Mastering the RAG Architecture: A Scientific Approach to Building Domain-Specific Chatbots

In today’s fast-paced Large Language Models (LLM) landscape, the Retrieval Augmented Generation (RAG) architecture emerges as a game-changer. RAG is a novel architecture that enables the use of LLMs like GPT-3.5/4 or LLAMA to build domain-centric chatbots without the need for expensive fine-tuning. It employs clever techniques to identify relevant contexts from the data, which can then be passed to the LLMs to synthesize answers. While it has been instrumental in several notable production use cases, including our own Eryl product under the GeneraX umbrella, the journey of RAG’s mainstream adoption is only just beginning.

At Affine, we don’t just adopt technology; we sculpt it. We have adopted a scientific approach to harness the capabilities of the RAG architecture for building production-grade LLM customer solutions.  This includes our Eryl product, showcasing the manifestation of our philosophy—implementing scientifically engineered solutions that resonate with individual customer requirements.

The RAG’s efficacy pivots around various design parameters. But how does one ensure peak performance? For us, it’s about a rigorous, scientific approach. We have borrowed significantly from the concept of hyperparameter tuning for Machine Learning and Deep Learning models. We systematically navigate these parameters, evaluating their performance on real-world test data – such as customer interactions in chat sessions that have received high Net Promoter Score (NPS) ratings, an industry-standard metric for customer satisfaction.

When it comes to building scalable, production-grade, and hallucination-free LLM applications, the key objectives are not only the accuracy of outputs but also factors like latency and cost of inferences. We evaluate the performance of all hyperparameters on all these factors and select or fine-tune iterations that rate high across all success factors.

Listed below are some RAG hyperparameters we utilize while developing LLM applications:

  1. Chunk Management Related: At the heart of RAG’s contextual retrieval lies a matrix of parameters – chunk size, overlap window, and top K chunks for retrieval (meaning the top K most relevant text chunks that are retrieved). Much like deep learning tuning, we employ an iterative but optimized methodology to discern the most effective combination.
  2. Embedding Model Fine-tuning: Fine-tuning the embedding model ensures the domain specificity of embeddings, thereby allowing retrieval of relevant chunks from the vector databases.
  3. Generator LLM Fine-tuning: By refining the synthesizer LLM on specific customer documents, it becomes attuned to unique nomenclatures and keywords. Given that this LLM steers the response synthesis, generating the final text that the end-users interact with, alignment with customer-specific lexicons is pivotal.
  4. Enhancement with Knowledge Graphs: Incorporating Knowledge Graphs with RAG becomes a force multiplier, especially for intricate, multi-contextual, or multi-hop queries, where the model needs to consider multiple factors or steps to generate an accurate response.
  5. Hard cutoff on Cosine Similarity: The conventional method of selecting Top K embeddings may still result in hallucinations, as for certain queries, none of the top K chunks may be relevant. In such cases, it is essential to have a hard cutoff on cosine similarity that only fetches chunks above the threshold.

Our approach involved systematically iterating through various combinations of the above design parameters in an optimized fashion and evaluating the performance on test data. It should be noted that iterations involving fine-tuning embeddings or generator LLM models can be computationally expensive and should be undertaken only if the development budget allows.

The following capture key performance metrics and other ML/LLM hygiene practices that we adopt in building the LLM application:

  1. Performance Metrics: Our benchmarking isn’t just about accuracy. By analyzing real human chat logs with high NPS scores, we gauge efficacy. Additionally, parameters like latency and cost of inferences help construct a system that’s precise, economical, and prompt.
  2. Optimization within Boundaries: Despite the computational complexity, especially when fine-tuning the embedding and generator models, we ensure that development remains within budget constraints, thus achieving a balance between performance and cost.
  3. Systematic Record-Keeping with MLOps: Tools like MLflow are invaluable, enabling us to meticulously document all iterations, providing a robust framework for tracking changes, and ensuring that the model can be easily deployed or rolled back as needed.

The culmination of these steps results in an LLM solution that’s not only primed for production but also accurate, cost-effective, and systematically built, ensuring reproducibility and reusability.

In summary, the RAG architecture isn’t merely an innovation in building QnA systems; it’s a game-changer in the realm of large language models. By enabling specialized chatbots to leverage the power of LLMs without the need for expensive fine-tuning, our Eryl product exemplifies how the intelligent use of LLMs, enabled by RAG, can yield a product that is not only cutting-edge but also finely tuned to meet distinct customer needs.

At Affine, we don’t merely adapt to technology; we shape it, refine it, and make it our own. We continually integrate groundbreaking technology into our ethos of delivering scientifically engineered solutions, creating products that are not just innovative but also tailor-made to tackle real-world business challenges head-on.

As we continue to advance in this journey, the RAG architecture stands as a cornerstone, showcasing the incredible potential and adaptability rooted in the synergy between retrieval and generation techniques in LLMs. We aim to go beyond just building chatbots; our vision is to build intelligent systems that can understand, learn, and adapt, setting new standards for what is achievable in the realm of artificial intelligence.

What is Web3? What are its Use Cases?

In recent years, we have witnessed a massive shift towards digitization across various industries, from finance to healthcare, education, and entertainment. Digital Transformation has brought numerous benefits, such as convenience, efficiency, and accessibility. However, it has also created new challenges, such as centralization, data breaches, and privacy concerns.

Here comes Web3! It’s a new generation of the internet that promises to address these challenges by leveraging the power of decentralized networks. In this blog, we will explore the exciting world of Web3 and its potential to revolutionize how we interact with cyberspace. So, buckle up and get ready to uncover the future of decentralized digitization with Web3!

What is Web3?

The current web we use to access and share information is the 2nd generation. In the 2nd version of the web, the content we produce is saved in a central server controlled by an authority. Various data ranging from emails, health tracker data, shopping interests, social media posts, photos, entertainment, and choices to web browsing patterns and other forms are the data collected on a regular basis from the user and saved under a centralized service provider storage where users have no control over their data.

The true ownership of this data has never been owned by the user but rather by the central authority controlling the service. Web3, which is the 3rd generation of the web, will solve this critical ownership problem by shifting the control of content from central authority back to the users. Users have complete control over what they share and with whom they share and can completely revoke the permissions at any time. Web3 is all about less trust and more truth.

How will Web3 be different from Web2?

The real necessity of Web3 – Let’s look at real-life use cases that have facilitated the design thinking towards web3:

Use Case 1:  Many of us have played or heard of the popular flash-based game called Farmville, which was designed by Zynga on Facebook. In 2020 after 11 years of service, the development has been ceased leaving millions of fans of the game unable to access the game assets they’ve purchased over the years. Web3 can solve this problem by transferring the ownership of those assets as limited-time collectibles to the fans who bought them on an open decentralized marketplace.

Use Case 2: The fundamental problem that occurred when the popular social media site Orkut got shut down, resulting in millions of users losing access to their photos and posts shared over the platform, which are actual memories from the early days of the web in the 2000s. Web3 can solve this problem by bringing back the control of user data (posts, media) to the users and freedom to take the data to their platform of choice by making it interoperable.

Use Case 3: Free speech is a powerful principle of democracy that should be censorship resistant. There are many cases of social media accounts getting banned just because of criticizing authority of its flaws even though when it’s the truth, which indicates the suppression of the free flow of open speech. Essentially the accounts have been permanently locked in their previous posts on social media. A web3 based existent decentralized social media platform like Mastadon solves this problem where users can control the data they publish and interoperate with other platforms of their choice where there should always be one single source of truth that is censorship resistant.

What are the benefits of providing access to user data?

Healthcare data, for instance, can be shared with various medical sources for advancements in medical research, where the data exchange will be peer-to-peer. Our photos & media, meanwhile, can be permitted to be uploaded to Facebook, Instagram, Flickr, etc., without uploading individually. And the most important aspect of any web3 application should be the incentive structure the user can benefit from companies accessing their data. Users by choosing and providing access to their data should be incentivized for the contribution, which is clearly lacking in the web2 world.

Is Web3 based on blockchain?

One of the misconceptions most people believe is that web3 is completely blockchain-based. But the truth is that web3 is a culmination of technologies, whereas blockchain is a mere part of web3. For instance, we imagine blockchain like Bitcoin/Ethereum provides a solid trustless, permissionless cross-border payment between individuals without any central banking authority to control the transaction. Blockchains are excellent use cases for web3 where public platforms like incentive structure, decentralized access, decentralized finance, NFTs, and DAOs can be built to support the principles of web3 ideology. Even standardized technologies can be part of a web3 application development, given it implements basic principles of user privacy, ownership, and censorship-resistant data flow.

Web3 and Gaming Applications

As we see a trend towards adaptation of web3, we will see more games built around incentivizing the users. Game designs will make use of releasing limited game assets as collectible NFTs to its fans, thereby making them a partner in the development process and creating a win-win scenario when the game performs well for both the companies and fans alike. Users can be assured that they will still own the game assets as collectibles even though the game shut down in the future.

Web3 and Defi (Decentralized Finance)

The true potential of Finance will be unlocked when more financial products are implemented around the principles of Web3 and Decentralized Finance. Already existing applications like Uniswap and Airswap have taken the first steps in the evolution of Web3 financial products. Imagine finance becoming peer-to-peer between any two parties in the world where the transaction rules are governed by a contract running on a trustless network autonomously. This removes a whole lot of unnecessary paperwork and intermediatory fees and, most importantly, saves a lot of time for instantaneously accessing various financial products, even in remote places of world where banking is a luxury. Decentralized cross-border payments are the future.

Web3 and Metaverse

The Metaverse is a digital platform that provides an immersive experience to users using AR and VR technologies. We can view this as a 3D web where users can have 3D interactions with other users, bots, and applications. Metaverse as a platform will be there for enhanced social connections. Imagine Facebook as a 2D place where you can add a friend, chat with someone, join a group, etc. The same actions can take place in Metaverse in 3D with enhanced user experience and social connections. Web3, in some ways, will be a component of this digital social experience by powering apps that are censorship resistant, decentralized, and secure.

Web3 and AI

Eventually, AI is the umbrella term where the full potential of Web3 principles comes into play. By owning the data in various forms, users will have complete control over who to give access to, thereby getting an incentive for doing so. Imagine companies building AI models having access to the same reliable and quality data from real users who are willing to participate in their development activity. The users have the right to control the information to share and get incentivized, and the companies have access to golden data to build better AI models which perform well than the ones trained on noisy data. Web3 principles will govern the flow and access of this data by creating a more inclusive environment.

Summing up!

Privacy by design and default, less trust and more truth, whereas decentralized and censorship-resistant ownership is one of the principles of any future Web3 application. An ecosystem where humans/bots/ devices/applications can securely operate on a trustless network can be enabled by following these principles. While Web3 is primarily a concept under development today, some early applications demonstrated its implementation, such as Odysee, a decentralized video-sharing app, and NFT marketplaces where users have the freedom to sell an NFT on a platform of their choice by just connecting their wallet, Mastadon Social Network, etc. In Web3, we can even imagine building decentralized machine learning models that can perform more efficiently.

Who Will Lead the Creator Economy: AI Or Creators?

Gone is the era of big media; we are now entering the pinnacle of the content-creator economy. The current creator economy stands at around $104.2 billion in market value. It has successfully bypassed traditional gatekeepers and established the connection between creators and viewers, be they influencers, artists, gamers, and so on. Various businesses, large and small, now collaborate with creators and online influencers to enhance their brand image and visibility amongst fans. As a result, influencer marketing has grown tremendously last six years

But we’re just scratching the surface of the potential in the creator economy. As with other industries, AI will lead the creator economy into the next era, where creators will flourish.

AI is at the Helm of the Creator Economy

One might argue that innovators and creators are at the center of the creator economy, but I see creators as the face, while AI is the science behind the creator ecosystem.

Sure, the creativity and innovation involved in creating content is the responsibility of the content creator, consider things from a business standpoint for a moment: the content has to reach the right audience, and the platform here is the internet.

Content creators agree with this too. According to research, the biggest challenge for content creators is getting their content discovered by the right audience.

YouTube, Instagram, and TikTok are the most popular Short Video Format (SVF) content platforms today and host millions of content creators, providing success to both parties involved—businesses and creators. As of 2020, 22,000 YouTube creators have more than one million subscribers, a 65% growth increase from 2019.

This signifies the democratization of content creation and viewership and the fall of the big media over the years as the internet picked up the pace. The robustness of AI technology has made it possible for small and niche content creators to shine and make a living out of content creation. 12% of full-time content creators earn more than $50k/year, while 9% of niche content creators earn more than $100k/year.

Content creation and content recommendation are the equivalents of demand and supply, and content recommendation just so happens to be the foreground of AI. User interests, location, preferences, and hobbies are evaluated to fuel their content consumption through practical content recommendations from various creators.

As for creators, there is a basic formula here. There are unique creators for sure, but there are also creators that have a common audience with a common subject theme for content. The identification of this theme and content type is all thanks to AI, which helps content creators generate the finest ideas for their audience. It also provides multiple inputs, i.e., areas of improvement in current content types, considering the most viewed videos by the creator and similar creators.

As I mentioned, a content discovery gap is a massive challenge for content creators, and AI is the best bet to close this gap.

Creator Economy has redefined Social Commerce

You are on a social platform and find an excellent product. But when you click to buy it, you are redirected to another app/website, which breaks immersion. Most mobile shoppers will agree that an intrusive shopping experience is a major turnoff. But with social commerce, this is history. Customers can now not leave the social platform and purchase a product on the platform, introducing a seamless shopping experience that has now become the gold standard. According to Deloitte research, social commerce powered by the creator economy is set to grow to $2 trillion by 2026.

The reason for this is quite simple: buyers and non-buyers consume a lot of content on social media platforms, which is a significant reason for the growth of platforms like TikTok, Instagram, and YouTube. The content creator economy only furthers such user behavior.

A majority of what was previously a retail experience has now been shifted to a social experience. Content creators on these platforms come in as brand influencers, driving marketing for brands and purchases, directing views, and leaving affiliate links to motivate followers. Content creators/influencers are then monetized by companies based on the purchases.

While customers benefit from a seamless shopping experience, companies can conduct transactions on a specific platform without drop-off rates while controlling the entire funnel the buyer goes through.

So, how has this worked so far?

The estimated CAGR growth from 2020 to 2025 shown above looks exceptional, and the 10-year prediction is easy to achieve with the right approach. The social commerce ecosystem diversifying rapidly due to accelerated technical advancements and millennials and Gen Z’s increased reliance on and adoption of social media. The companies need the right approach: leveraging content creators/influencers on the respective platforms suitable for their products/services to influence the audience and their buying behavior.

Brand Partnership-Influencer Marketing: Monetization in the Creator Economy

Platforms like YouTube have been functional for a long time, but the creator economy is fairly recent. One of the primary reasons for the drastic shift towards an online creator economy is social media. AOL, MySpace, and Facebook were initially platforms to connect and find new friends, but the social media explosion was a phenomenon that had more in store, especially for businesses.

The continued use of social media expanded its horizons and gradually resulted in new buyer-driven trends, where people turned to social media platforms for shopping. I remember a time when having a website was the smart ‘business’ thing one could do, but today a social media presence is mandatory if you wish to acquire new users for your business.

39% of customers used Facebook for online purchases, while 29% used Instagram.

And these platforms had their icons- the content creators and influencers with the audience. Brands selling online knew they got closer to their customer with these influencers; even better, there are potential buyers amongst the influencer audience.

With content creators, partnerships with brands aren’t subject to the typical norms. Creators are infusing innovation into promoting the brand’s offerings, and the audience feels a genuine connection towards their favorite content creator. This is where brands can use AI to leverage and reach the right audience using influencers, content, customer behavior, and critical buyer data.

Even most content creators majorly depend on brand collaborations for their income. It’s a win-win for both the brand and the content creator.

AI is at the core of the reward system for creators while improving brand visibility, influencing users toward the purchase, and improving the brand’s ROI. And just like that, we are ushering in a new era of the content creator economy with a democratized approach.

How Exactly Does AI Help Here?

For starters, content recommendation engines suggest creator content to the relevant target group, making the reach more effective, choosing quantity over quality. The audience gets recommended the content they want to consume and not just something that is popular despite being irrelevant to them.

Brands have great options here. They can choose creators apt for their offerings among the millions, leveraging AI, considering their target audience, consumer behavior, previous purchases, and social media footprint.

The content creator has its fair share of problems, but as with other industries, it is one with great potential. Not all creators are equally rewarded, and not all get the credit they deserve. Niche content creators see fair success, and the content consumption landscape is shifting from conventional quantity to new quality. It is a win for both creators and consumers, but with the application and leveraging of AI, the biggest victors will be businesses.

What Does Affine Bring to The Table?

Affine is a pioneer and a veteran in the data analytics industry and has worked with giants like Warner Bros. Theatricals, Zee 5, Disney Studios, Sony, Epic, and many other marquee organizations. From game analytics, media, and entertainment to travel & tourism, Affine has been instrumental in the success stories of numerous Fortune 500 global organizations; and is an expert in personalization science with its prowess in AI & ML.

Learn more about how Affine can revamp your online business!

The Rise of On-Demand Industry: A Case of AI Marvel

With a drastic shift in consumer behavior and day-to-day lifestyle (work from anywhere/work from home), the way people interact and purchase from businesses has evolved. One industry, in particular, has leveraged this to the fullest. On-demand services have seen unreal growth in recent times.

By 2025 the market demand of the on-demand app economy is set to hit $335 Billion.

The storefront barrier has been removed, and this has opened up massive opportunities for customers to get services at their place, be it ordering food, and groceries, watching a movie, or even booking therapists, home maintenance, and repair services. Businesses have tasted success too. Doordash, Grubhub, and UberEats make up 96% of the food delivery market in the USA!

All the on-demand services/apps have an underlying element powering them- the robust tech of AI, ML & Cloud.

In fact, these technologies are singlehandedly responsible for making the on-demand industry lucrative, offering ease of access and immense benefits to the customers at a convenient tap of their device screen.

Customer Service is the ace card for the On-demand industry.

With the explosion of the internet and social media, I’ve observed a paradigm in human behavior. Two things have significantly gained traction-

  1. Instant gratification
  2. Low attention span

The on-demand industry thrives on these behaviors. Grocery, food, and home services are availed within minutes of booking from the user. Customers love this phenomenon, and this is one of the major reasons the on-demand industry has blown up significantly.

But there is a last-mile gap that still needs to be perfected-Customer Service. Customer Service remains the last piece of the puzzle for on-demand businesses.

In my opinion, the most obvious solution is personalization. From a customer angle, the demand is quite simple. They want what they are looking for, and they want it fast. The customer isn’t ready to wait or compromise on any other front.

The business challenge

From the business side of things, meeting this demand is easier said than done. Personalization is the key to offering the best possible service and user experience. Businesses need to predict user demand and deliver accordingly and exactly what the customer wants.

  • Data is large and unorganized. Making sense of it is no easy task, and it’s easy to get lost in the translation
  • A large number of online users spread across demographics means customer service is challenging
  • Personalization isn’t as simple as using a complex tool. The marriage of the right data and the right tool with efficient targeting is the only way for effective reach

Businesses do believe that they provide superior customer service, while customers vehemently disagree.

To offer the best personalization experiences on-demand businesses need specific user data and the right tool to leverage it to reach out to the customers.

Location-based data, user interests, and preferences combined with historical purchase data act as the base ingredient that is a necessity to offer excellent customer service through hyper-personalization made possible with AI.

The solution

Marketing Campaign Personalization Engine is an Affine’s AI-powered solution that helps on-demand businesses personalize product/service offers and customer messaging most likely to get engagement from their users. Behavioral insights and personal data are combined with machine learning to produce the best possible personalization output to reach, engage and provide the best customer experience while optimizing promotional costs and efforts through smart automation.

High-value customers, loyalty, and repeat business

There are occasional customers, and then there are repeat, loyal customers. Customer acquisition is a crucial part of any business, and so is the case for On-demand businesses.

I’ve seen signup offers, 60% off for new customers, and many lucrative deals that are focused on bringing new customers. However, acquiring a new customer is up to 25 times more expensive than retaining existing customers.

While customer acquisition efforts must go and is a basic aspect of any on-demand business, customer retention is equally, if not more important, to boost revenue. 65% of an organization’s business is from existing customers!

They’ve already purchased from your business; it is just about retargeting and pushing them to do it again. But there’s an even better potential here for on-demand businesses. That is targeting high-value customers.

The challenge

  • All on-demand businesses have loyal customers as well as high-value customers, but many businesses have a tough time identifying them
  • The challenge is obtaining valuable customer data to identify these high-value customers
  • Many businesses, without rightly identifying high-value customers, end up targeting generically and spending quite on marketing budget while not making revenue to justify it

The solution

Affine’s Customer Loyalty Analytics solution helps the on-demand business ecosystem to extract more out of targeting and retargeting exercises.

The AI-driven solution analyzes vital user data like purchase history, purchased items, and the frequency of purchases/orders and creates a detailed customer profile. This way, it helps on-demand businesses segment valuable customers and even tiers of customers with exclusive benefits.

  • This makes targeting and retargeting an immensely efficient procedure, optimizing the promotional spend
  • A huge amount of data is given a structure that is actually usable for the businesses and adds value to their targeting efforts
  • By creating tiers for the customers, offers can be better personalized, and cross-selling will result in more spending per order from these customers, increasing business ROI

On-demand is not a bubble anymore; the future looks promising!

From food delivery to healthcare, the on-demand service industry isn’t a fluke by any means; it is set to grow exponentially. On-demand users spend up to $57.6 Billion a year, and the numbers will only go up with new players offering better services debuting in the market every day.

There’s no need to reinvent the wheel here. Pick any established service from the on-demand industry. There’s always room for improvement. By offering better customer experience, personalization offers, and an overall shopping experience, it can penetrate and sustain the lucrative on-demand industry. Even traditional industries will see a shift towards the on-demand economy in the future, so the potential is ripe!

But that’s the catch. You need the right tools for the job, and in this case, AI is the only tool that can provide a long-term growth scenario for any player in the on-demand industry. AI is the backbone of the on-demand industry and, paired with data, can work wonders in terms of delivering exceptional customer experience while streamlining and improving business productivity, boosting revenue in the long run.

What does Affine bring to the table?

Affine is a pioneer and a veteran in the data analytics industry and has worked with giants like Warner Bros Theatricals, Zee 5, Disney Studios, Sony, Epic, and many other marquee organizations. From game analytics and media and entertainment to travel & tourism, Affine has been instrumental in the success stories of many Fortune 500 global organizations; and is an expert in personalization science with its prowess in AI & ML.

Learn more about how Affine can revamp your On-demand Business!

AI in healthcare: What awaits the future of the industry?

AI is known to perform tasks faster than humans while being more efficient. The application of AI in healthcare is growing into a phenomenon to look forward to as it is making waves across the healthcare industry.

The use of AI in healthcare can help ease the lives of doctors, patients, and healthcare providers alike. These are quality-of-life additions to the healthcare industry and are highly capable of predicting trends using analytics and data, leading to medical research and finding innovation. Combined with patient data and predictive algorithms, it is now possible to identify cancer in its nascent stages and heart attacks beforehand.

Robotics with AI isleveraged for surgery owing to their superior precision and can even perform actions best suited for the medical scenario.

On the administrative front, intelligent healthcare solutions bring a lot of structure and efficiency to day-to-day operations for healthcare companies.

The global AI in healthcare will see a CAGR of 37.1% from 2022-2023!

So, it’s evident that AI in healthcare is brimming with potential. Here are some of the benefits of AI in healthcare:

  • Efficiency in clinical research, diagnosis, and treatment
  • Precision in advanced surgeries
  • Optimizing costs for healthcare organizations
  • Streamlining processes and administration for healthcare companies
  • Improving patient-doctor and patient-healthcare provider experiences

Contrary to popular opinion, using AI in healthcare will not replace humans but will make their lives easier. It will facilitate an optimal team effort scenario, streamlining the process and obtaining maximum efficiency, which is favorable for healthcare providers and patients.

Centralize Patient Information and Streamline Operations with AI in Healthcare

The accelerated shift towards the online ecosystem on all fronts is now a reality. It’s not just eCommerce businesses; even healthcare companies require an online presence nowadays to capitalize on patients. Healthcare providers use advanced medical devices, and with the current norm of connected technology, there is an inflow of an immense amount of data.

The Bigger Challenge?

While some organizations have solutions to manage data and generate patient information, the challenge lies in bringing structure to the data.

Healthcare organizations are left with a plethora of unstructured data, sometimes from multiple sources. Even top healthcare companies lack the proper means and solutions to manage data ranging from patient information and medical transcripts to medical records. Efficiently streamlining data from various information sources is a significant challengeandcannot be performed with relational databases.

The Universal Customer Data Platform is Affine’s AI solution for healthcare providers that combines multiple data sources and pools the information, creating a universal data lake.

Healthcare providers can leverage insights from this and create a universal patient profile that can be accessed worldwide. They can also personalize their efforts for customer reach and engagement more efficiently, owing to the centralized data ecosystem.

Ai In Healthcare to Optimize Costs And Increase Revenue

Healthcare prices stem from various factors. The United States is notorious for its ridiculously high healthcare costs, which spells a significant setback in terms of revenues for healthcare providers. The USA spent a whopping $4,124 Billion on healthcare in 2020! Modern-day inflation has made everything high-priced and expensive healthcare is a deterrent to many patients who ideally should take regular tests and visit practitioners. In the long run, this is an unsustainable model for healthcare companies.

Healthcare providers can create packages and plans that patients can subscribe to, ensuring business while providing value to patients both service and cost-wise for a sustainable operating model in the long run.

The Intelligent Pricing Solution is another Affine solution for healthcare providers that creates a uniform pricing model for patients. This smart solution only recommends required tests and procedures thanks to the patient information, thus making it possible for the companies to offer great prices while providing quality healthcare service.This is vital in developing a sustainable patient relationship while improving engagement and helps increase the ROI in the long run.

Use Of AI In Healthcare Is Essential for A Sustainable Future

The above solutions are just a glimpse of what AI is capable of for healthcare providers. According to this report, ML in the healthcare industry is estimated to generate a global value of over $100 Billion. In the future, we are looking at AI-powered predictive healthcare, where anomalies in human health and chronic diseases are identified with historical patient information, and preventive measures can be taken in advance to save lives.

People are embracing travel post the pandemic, and organizations are encouraging a work-from-anywhere culture. Healthcare needs a connected tech approach, and providers must opt for centralized AI solutions that can provide patient information oncommand. Not only does this make for an efficient, but it also improves the patient experience.

For healthcare providers, AI and ML-based solutions help engage and motivate patients, improve their lives, assist them in their day-to-day activities, and handle the inflow of customers smoothly.

This may sound quite simple, but we all witnessed what happened when COVID-19 hit the USA, and the healthcare system got choked handling the sheer number of patients. Despite having one of the most advanced healthcare systems in the world, the country was brought to its knees during the pandemic. When I mentioned earlier that AI assists humans and does not take over their jobs, I couldn’t find a better example than the healthcare industry. Healthcare employees are among most overworked, burnt out, and underrated.

AI in healthcare can aid organizations, patients, and healthcare employees at capacity and their absolute limit, helping them bypass various tasks.

What does Affine bring to the table?

Affine is a pioneer and a veteran in the data analytics industry and has worked with giants like BSN Medical, Optum, AIG, New York Life, and many other marquee organizations. From game analytics and media and entertainment to travel and tourism, Affine has been instrumental in the success stories of many Fortune 500 global organizations; and is an expert in personalization science with its prowess in AI & ML.

Learn more about how Affine can revamp your healthcare business!

Decision Intelligence: The Next Big Milestone in Impactful AI

As businesses take a global route to growth, two things happen. First, the complexity and unpredictability of business operations increase manifold. Second, organizations find themselves collecting more and more data – predicted to be up to 50% more by 2025. These trends have led businesses to look at Artificial Intelligence as a key contributor to business success.

Despite investing in AI, top managers sometimes struggle to achieve a key benefit – enabling them to make critical and far-sighted decisions that will help their businesses grow. In an era of uncertainty, traditional models cannot capture unpredictable factors. But, by applying machine learning algorithms to decision-making processes, Decision Intelligence helps create strong decision-making models that are applicable to a large variety of business processes and functions.

The limitation of traditional AI models in delivering accurate decision-making results is that they are designed to fit the data that the business already has. This bottom-up process leads to data scientists concentrating more on data-related problems rather than focusing on business outcomes. Little wonder then that, despite an average of $75 million being spent by Fortune 500 companies on AI initiatives, just 26% of them are actually put into regular use.

Decision Intelligence models work on a contrarian approach to traditional ones. They operate with business outcomes in mind – not the data available. Decision Intelligence combines ML, AI, and Natural Language queries to make outcomes more comprehensive and effective. By adopting an outcome-based approach, prescriptive and descriptive solutions can be built that derive the most value from AI. When the entire decision-making process is driven by these Decision Intelligence models, the commercial benefits are realized by every part of the organization.

Decision Intelligence Delivers Enterprise-Wide Benefits

Incorporating Decision Intelligence into your operations delivers benefits that are felt by every part of your business. These benefits include:

  1. Faster Decision-Making:
    Almost every decision has multiple stakeholders. By making all factors transparently available, all the concerned parties have access to all the available data and predicted outcomes, making decision-making quicker and more accurate.
  2. Data-Driven Decisions Eliminate Biases:
    Every human process data differently. When misread, these biases can impact decisions and lead to false assumptions. Using Decision Intelligence models, outcomes can be predicted based on all the data that a business has, eliminating the chance of human error.
  3. Solving Multiple Problems:
    Problems, as they say, never come in one. Similarly, decisions taken by one part of your operations have a cascading effect on other departments or markets. Decision Intelligence uses complex algorithms that highlight how decisions affect outcomes, giving you optimum choices that solve problems in a holistic, enterprise-wide way, keeping growth and objectives in mind.

Decision Intelligence: One Technology, Many Use Cases

Decision Intelligence tools are effective across a multitude of business applications and industry sectors. Here are some examples of how various industries are using Decision Intelligence to power their growth strategies:

  1. Optimizing Sales:
    Decision Intelligence can get the most out of your sales teams. By identifying data on prospects, markets, and potential risks, Decision Intelligence can help them focus on priority customers, predict sales trends, and enable them to forecast sales to a high degree of accuracy.
  2. Improving customer satisfaction:
    Decision Intelligence-based recommendation engines use context to make customer purchases easier. By linking their purchases with historical data, these models can intuitively offer customers more choices and encourage them to purchase more per visit, thus increasing their lifetime value.
  3. Making pricing decisions agile:
    Transaction-heavy industries need agility in pricing. Automated Decision Intelligence tools can predictively recognize trends and adjust pricing based on data thresholds to ensure that your business sells the most at the best price, maximizing its profitability.
  4. Identifying talent:
    HR teams can benefit from Decision Intelligence at the hiring and evaluation stages by correlating skills, abilities, and experience with performance benchmarks. This, in turn, helps them make informed decisions with a high degree of transparency, maximising employee satisfaction and productivity.
  5. Making retail management efficient:
    With multiple products, SKUs and regional peculiarities, retail operations are complex. Data Intelligence uses real-time information from stores to ensure that stocking and branding decisions can be made quickly and accurately.

Incorporating Decision Intelligence into the Solutions Architecture

CTOs and solutions architects need to keep four critical things in mind when incorporating a Decision Intelligence into their existing infrastructure:

  1. Focus on objectives:
    Forget the data available for a bit. Instead, finalize a business objective and stick to it. Visualize short sprints with end-user satisfaction in mind and see if the solution delivers the objective. This approach helps technical teams change their way of thinking to an objective-driven one.
  2. Visualize future integration:
    By focusing on objectives, solution architects need to keep the solution open to the possibility of new data sets arising in the future. By keeping the solution simple and ready to integrate new data as it comes in, your Data Intelligence platform becomes future-proof and ready to deliver answers to any new business opportunity or problem that may come along.
  3. Keep it agile:
    As a follow-up to the above point, the solution needs to have flexibility built in. As business needs change, the solution should be open enough to accommodate them. This needs flexible models with as few fixed rules as possible.
  4. Think global:
    Decision Intelligence doesn’t work in silos. Any effective Decision Intelligence model should factor in the ripple effect that a decision – macro or micro – has on your entire enterprise. By tracking dependencies, the solution should be able to learn and adapt to new circumstances arising anywhere where your business operates.

Machine learning and artificial intelligence are niche technologies, and companies have started thinking about or utilizing these technologies aggressively as part of their digital transformation journey. These advancements have changed the demand curve for data scientists, machine learning, and artificial intelligence technologists. Artificial intelligence-driven digital solutions require cross-collaboration between engineers, architects, and data scientists, and this is where a new framework, “AI for you, me, and everyone,” has been introduced.

To Sum Up

Decision Intelligence is a powerful means for modern businesses to take their Artificial Intelligence journey to the next level. When used judiciously, it helps you make accurate, future-proof decisions and maximize customer and employee satisfaction, letting you achieve your business objectives with the least margin of error.

E-Commerce, AI & the game of customer experience

Exceptional customer experience through personalization is the elixir of online shopping. 95% of companies increased their ROI by 3X with personalization efforts. There is always the debate about privacy and handling of data, but the younger next-generation of shoppers are open to it, and the future is set to see data-backed hyper-personalization across multiple touch points in E-Commerce platforms, in my opinion.

The challenge for E-Commerce businesses with personalization is multifold. Customers seek a device-agnostic experience in online shopping, so a uniform shopping experience is a basic need. On top of that, there is a plethora of data inflow, which makes for great potential but only if it is leveraged aptly by E-Commerce businesses. With the prowess of AI, the possibilities are many and colorful for E-Commerce businesses.

There is an immediate need for a centralized solution, considering the omnichannel nature of shopping amongst online customers. With data inflow from various sources, the alternative is just a large amount of unstructured pile of data that does not provide value for the business.

This is where AI will rewrite the rules and set the foundation for a new type of unparalleled customer experience in E-Commerce.

Data & AI-2 sides of the personalization coin for E-Commerce

Personal opinion -I’m willing to share data for a no-nonsense experience on the internet. Eventually, with connected tech and synced accounts becoming a norm by the day, online shoppers will opt for the same, considering the benefits. In my opinion, we’re too co-dependent on the internet for our day-to-day activities and sharing data for a better shopping experience is a valid tradeoff.

3d product previews & virtual try-on are the tips of the iceberg of what AI has in store to revolutionize the customer experience and write the future for E-Commerce companies.

AI in movies is mostly dystopian, about rogue software taking over the world, but in the real world, there is a drastic contrast to the applications and potentials of AI in day-to-day use cases. AI isn’t anymore a niche tech reserved for a few organizations.The infusion of technology in E-Commerce has made life easier for people, and the data shows. Just take a look at the worldwide sales, the growth is unprecedented and projected to spike through the roof in the coming years.

So, naturally, as someone working in close quarters with AI experts and E-Commerce veterans, I couldn’t help but speculate about the future of the marriage between the two.

Next level personalization for exceptional user experience

I remember the good old days of AOL messenger, Yahoo-Mail & My-Space. The internet at its cusp was pure, and people just hung out and wondered at the marvel of this infant tech and its dynamic capability.

With great power comes great responsibilities, and the internet opened the gateway to access various types of data; this, paired with the eventual shift towards online shopping, would work wonders for businesses. This was inevitable as the monetization of the internet was bound to happen. When anything is free, you’re the product.

I’m spoiled by the Apple ecosystem and can’t get enough of the seamless interaction and experience when I use any device, be it my laptop, mobile, or iPad. Imagine this for E-Commerce businesses on a gigantic scale.

For E-commerce companies, AI technologies can vastly improve product selection and overall user experience, which becomes a crucial part of the customer lifecycle.

Big data offers a plethora of opportunities for E-commerce personalization. Segmenting and customizing the experience is made possible by analyzing historical data. Even the messaging is personalized to the core, ensuring the relevancy factor is preserved.

Affine’s AI-powered Hyper Personalized Marketing Experience is one such solution that leverages customer behavior data, segmentation data, and cohort analysis at scale

What do customers gain from this?

As an avid online shopper, I appreciate the personalized shopping experience on an E-Commerce platform. Getting to see the products with utmost relevance is a positive shopping experience rather than being spammed with random recommendations; at least, that has been my experience so far.

When E-Commerce businesses leverage data properly, it creates an exceptional shopping and customer experience and results in return customers. We all have our favorite shopping websites, and personalization has played a crucial role in them.

What do E-Commerce businesses gain from this?

First things first: – a centralized workhorse robust personalization ecosystem,minimal data leak, and making sense of every customer move and converting it into a quality-of-life add-on with personalization, targeting, and segmentation.

Such a sophisticated solution makes it possible to identify the most feasible customers and target them with hyper-personalized efforts. The data is the only limitation to the possibilities here. Not only does this net the best possible customers for the business, but it optimizes the marketing efforts making the whole targeting process efficient. Ad budgets are kept in check while ensuring superior ad display and performance.

I’ve spent fat budgets on marketing only to receive lukewarm prospects. I’ve also spent conservatively with proper data-powered personalized marketing initiatives that have resulted in excellent conversions. Centralized AI solutions provide complete accountability for marketing spend, helping E-Commerce businesses measure every dollar spent per lead.

Automation is seamless with AI

Contrary to popular opinion, AI is not about taking over human jobs; it is more about making our lives easier. Customer support is one of the most tasking efforts in the E-Commerce business. I’ve had my fair share of endless calls with customer support with language and technical issues. I know you’ve all been a part of at least one such instance in your lifetime.

The potential for customer support automation via chatbots is a fast-tracking of the entire process and is already seen with giants like Amazon.

E-Commerce businesses, small or large, can leverage AI to automate mundane tasks and basically run your online store to an extent without human intervention and take care of the overall processes in their day-to-day activities.

Be it automating marketing efforts, support, customer engagement, or advertising, AI has the clout to not only handle but also excel in outperforming manual labor while casually being time and cost-efficient in the process. Automating the marketing process also brings up a 451% increase in qualified leads.

Combat fraud activities with AI’s security prowess

We’ve all known of someone who has fallen prey to online fraud and I know I do. With the convenience of seamless transactions, there is also the vulnerability of fraud in E-Commerce transactions. A research shows that online payment fraud could lead to a global merchant loss of up to $343 billion from 2023 to 2027.

Despite security measures, these instances call for drastic innovations in verification tools and authentication processes. AI and ML algorithms are the only effective option to combat online threats at such a higher scale in the count of millions of transactions, with predictive analytics to detect behavioral patterns and suspect behavior resulting in fraudulent transactions.

Brand loyalty, repeat business, and ROI

A gazillion data points are analyzed so that customers can get the best shopping experience and keep coming back for more. Sure, but E-Commerce businesses have the most to leverage from intelligent AI algorithms running through a plethora of data points. As I mentioned earlier, the possibilities are immense and lucrative.

I have my go-to sites for shopping, and I have valid reasons for that. So do customers, hence the need for AI in E-Commerce platforms. Personalization may make it sound like the customer is reaping all the benefits, but that’s the intention.

80 % of users saw an increase in lead generation, whereas 77% saw increased conversions when using automation.

The spend more get more is outdated, and efficiency is the need of the hour. Inefficient business operations cause a major dent in business revenues


The era of big data is here. Digital dominance is set to take over the world, and E-Commerce sites are more or less the new convenience stores in the digital age. I see a lot of competition; I see a lot of potential.

75% of businesses already use AI and are on the right track in the long run. Having run a startup myself, I understand the qualms about the cost of implementing AI. But the big picture justifies it and how.

AI in E-Commerce is not even a novelty anymore; it is a must-have. An online E-Commerce is a global marketplace, and the standards are already set. It’s excel or be expelled. A proper flow to guide your customers through their buying experience is the basic requirement to sustain and perform in the long run.

What does Affine bring to the table?

Affine is a pioneer and a veteran in the data analytics industry and has worked with giants like Warner Bros Theatricals, Zee 5, Disney Studios, Sony, Epic, and many other marquee organizations. From game analytics and media and entertainment to travel & tourism, Affine has been instrumental in the success stories of many Fortune 500 global organizations; and is an expert in personalization science with its prowess in AI & ML.

Learn more about how Affine can revamp your E-Commerce business!

Modern Data Platform vs. Traditional Data Platform: Where to invest your time and money?

Affine’s Analytics Engineering Practices releases Episode 3rd of Modern Data Platform, unbinding muddles between traditional and modern data platforms.

We explored data-driven strategy and the characteristics of modern enterprises in previous episodes of the Modern Data Platform series. Let’s take a closer look at some contemporary data-driven difficulties and see how traditional and modern techniques compared.

What are the bottlenecks of Traditional Data Platform?

Traditional data platforms fall behind the cloud on various factors. Organizations traditionally relied on in-house infrastructure to drive innovation and manage their workflow. Businesses would manage their in-house infrastructure and be accountable for everything, following traditional data platforms. The administration was taken care of by the in-house IT professionals. It was the business’s responsibility for downtime and repairs. We can say that managing the traditional data platform used to be an expensive affair.

Businesses had to take various responsibilities, like planning, people, hardware, software, and the environment. From a scalability perspective, it was possible, but it comes with the price of various challenges and delays hindering the performance of the enterprise’s overall data management.

Fig. 1: Illustrating the disadvantages of traditional data platform

Why should you consider cloud as a substitute option for Traditional Data Platform?

For instance, the traditional methods are still working effectively for many businesses, with a certain level of complications and challenges. Perhaps, the effectiveness gradually decreases as the business landscape changes with technology disruption.  The modern approach (cloud data platform) has a unique pace and benefits that can accommodate many inconsistencies at a time. The significant differences between the traditional and modern data platforms are highlighted in the table below:

How can Cloud Solutions boost your Modern Data Platform (MDP) Journey?

The cloud data platform is easing the use of data and securing it for every business. It provides various components and services like databases, software capabilities, applications, etc., that are engineered to leverage the power of cloud resources to decipher business problems. What are the benefits that businesses can achieve using Modern Data Platforms?  There are numerous benefits of MDP. Quicker time to develop and storage cost are a couple of crucial factors. Let’s take a look at other factors that comes into consideration.


It allows users to rent virtual computers to run their applications. IaaS has infrastructure and hardware running in the cloud. PaaS has application platforms and databases in the cloud. Containers are isolated environments for running software, and serverless functions compute services to run code in response to events.


A database service built and accessible through the cloud platform. It hosts databases without having to buy dedicated hardware. It can be managed by the user or as a service operated by a service provider. It can support relational and NoSQL databases. The database can be accessed from- anywhere through a web interface or an API.


It allows users to store and access data over a network, typically the internet. The cloud data platform offers flexible, pay-as-you-go pricing. Scalable up/down in near real-time. It supports backup, disaster recovery, collaboration/file sharing, archiving, primary data storage, and near-line storage.


Clients seek services that isolate resources, protect internet-facing workloads, and encrypt data on the network. Other networking aspects of a cloud platform include load balancing, content delivery networks, and domain name services.

Comparison of Cloud Service Providers

The Modern Data Platform is a future-proof architecture for business analytics. It is a functional architecture with all the components to support Modern Data Warehousing, Machine Learning, AI development, real-time data ingesting and processing, etc., and to leverage this, businesses need Cloud service providers who could provide services ranging from complete application development platforms to servers, storage, and virtual desktops. Based on the requirements, these providers are shortlisted to maximize the benefit to the enterprise. The cloud data platform helps enterprises extract high-value results to serve their customers in a data-driven future.

Affine’s Analytics Engineering practices can help with this seamless and effective transformation. Are you ready to embark on your journey to true data-centricity? We are here to help. Setup a call now!

The Modern Data Platform itself has proved as an asset to enterprises. So, what kind of data strategy and roadmap is needed to execute cloud-based solutions?

Well, that is the story for the next episode of the Modern Data Platform Series.

Till then, keep reading:

Episode 1: What Is Modern Data Platform? Why Do Enterprises Need It? – Affine

Episode 2: What are legacy systems? How Can Modern Data Platforms Bring Revolutionary Change? – Affine

About The Author(s):

The Modern Data Platform blog series is a part of research efforts conducted by the AE-Practices, which exists solely for hyper-innovation in the Analytics Engineering space. 

Our AE Practices is a dedicated in-house team that caters to all R&D needs. The team is responsible for continuously working on research tools and technologies that are new, happening, futuristic, and cater to business problems. We build solutions that are both cost and performance effective. Affine has always been deeply invested in research to create innovative solutions. We believe in delivering excellence and innovation, driven by a dedicated team of researchers with industry experience and technology expertise. For more details, contact us today!

What is Merge? Deciphering the future of the Crypto World

Cryptocurrency mining is extremely energy intensive and considered an environmental menace. Until recently, Ethereum was just another power-hungry cryptocurrency, but The Merge just fixed that.

With The Merge, the Ethereum blockchain has moved its consensus from proof-of-work to proof-of-stake and will burn significantly less ether (up to 99.95% less carbon footprint) compared to the previous proof-of-work method which used to burn billions of dollars’ worth of ether per year. I’m aboard the train of a greener earth and becoming more eco-friendly with sustainability while saving natural resources, which The Merge has helped further.

What is Merge?

The Merge is an upgrade of the Ethereum blockchain from a proof-of-work to a proof-of-stake system for authenticating new transactions. The new system replaces the old power-guzzling system.

With the Merge, the crypto industry sees an immediate slew of advantages:

  • Remains a decentralized platform
  • Extremely secure transactions
  • Up to 99% power efficiency compared to proof-of-work currencies like Bitcoin
  • Eliminates the need for miners and mining farms to authenticate transactions

Why is ‘The Merge’ a Marvel?

The open-sourced Ethereum is the host of DeFi protocols, NFTs, and cryptocurrencies valued at over a hundred billions dollars. All these digital assets were at stake, possibly being wiped out or irrevocably broken. The blockchain system verifies and processes new crypto transactions and migrating to a new type of system on the go is no simple feat by any means! It is akin to replacing an engine midflight without hiccups, but all it took for the merge to unfold was 15 minutes! I’ve been unwilling to participate in many Windows updates, which took more time, and I couldn’t use my device during the process!

It was around 2018 that I played around with cryptocurrency and even made a small profit. While I know many people who still gamble in the crypto market, I’ve since chosen to be a curious spectator of the volatile cryptocurrency market.  Many cryptocurrencies have debuted and seen their demise rapidly, but the two giants that have flourished so far are Bitcoin and Ethereum.

With The Merge, Ethereum is Now Environmental Friendly

It is a well-known fact that cryptocurrencies are power guzzlers.

I vividly remember the GPU hoarding craze recently and the large server farms used for Ethereum mining. The sheer power required to mine cryptocurrency is astounding and can power small-sized European countries!

This revolutionary feat made Ethereum 100x more efficient than before and that, which is nothing to scoff at. The long-term gains and the operational efficiency are remarkable and have left me awe-struck.

For all the environmental aficionados, The Merge is more of a festival thanks to the decrease in Ethereum’s footprint. However, that is more of a moral victory in anyone’s book

What Did This Mean for Investors and The Industry?

The Merge resulted in an interesting reaction from miners. I was excited about the energy efficiency and have been glued to crypto news, only to find that miners have moved on to greener pastures mining other currencies. The Merge has not created any substantial value for Ethereum as of now. Last I checked, Ethereum stood at a market capitalization of over $164 billion after the Merge. It used to hover over $200 billion.

The most significant advantage is that security will increase thanks to the proof-of-stake, which opens the option for varied granular incentives, unlike the proof-of-work consensus. This is because each validator’s stake is accessible.

I’m interested in seeing what types of rewards the stakeholders will earn. In the new consensus, validators are called stakeholders (people who stake a minimum of 32 Ethereum) in a secure network where trading is prohibited. In simple words, users earn Ethers by locking up their coins and validating transactions. It is impossible to predict the staking yields as more than 13.7 million ETH have been locked in the staking contract. stakeholders won’t be able to withdraw until a mid-2023 update. This should bring a little more structural stability and safety to Ethereum trading.

This is vital as the reputation of Cryptocurrencies in the past years has been tarnished thanks to hacks, scams, and billions of dollars lost due to new crypto companies shutting shop. I know a few people who bet big on Dogecoins, who thought it was the next Bitcoin but were left dejected.

The Merge Makes Ethereum More Secure Against Attacks

A significant upgrade I see is that the amount of Ethereum staked instead of resources spent defines network control. With the proof-of-work method, groups with large server farms could join forces to attack the network and sabotage others’ chances to update the ledger, opening options for higher rewards. With the new consensus, anyone trying to do that would be punished.

Although it seems like some people in the industry are a bit more apprehensive about the development. I was surprised by Changpeng Zhao, Binance’s CEO’s take on the Merge. According to him, the drop in Ethereum transaction fees is still a long-term reality, as he doesn’t expect it to fall drastically overnight. That seems to be the popular opinion, but the story might change in the long run.

Ethereum’s biggest competitor, Bitcoin, has no plans for such a migration. It is still energy-intensive and continues to run on the proven proof-of-work consensus. In fact, it was Bitcoin that pioneered the consensus, and Ethereum just followed suit. But competitors are not Ethereum’s concern now when enemies are in their camp.

Some Ethereum miners against the new consensus are fighting to keep the proof-of-work system alive. Their future is anyone’s guess as it’s impossible to determine the future token value.

Personally, other than the environmental-friendly badge, I don’t see any short-term advantages of The Merge.

What’s next for crypto?

The heavy carbon footprint was the least of the issues with Crypto’s bad reputation.

In my opinion, one of the biggest threats to Crypto has been government regulations, thanks to the various crypto scams. An imminent ransomware threat, major scams and a lack of proper understanding of the technology means the government will always be apprehensive about regulating cryptocurrencies. The Security and Exchange Commission’s recent claim that all Ethereum transactions belong to them raises many questions. Teething troubles post- The Merge, or any such innovation, are not surprising, but the division in the crypto community and the after-effects remain to be seen in the long run.

With the eco-friendly badge, I’m hoping the government regulations become a little more lenient and have a more explicit stance on cryptocurrencies. A clearer legal picture will give cryptocurrencies the much-needed relief and attract more people to invest in them. If eco-friendly cryptocurrencies are favored more by the government, we may see others jump ship, but I don’t speculate that happening anytime soon.

With The Merge, we only have bragging rights about the eco-friendly aspect. However, that will possibly soften governments’ legal stance on cryptocurrencies worldwide.

Don’t get me wrong, the Merge, to me, is a technical feat and is the right step towards a greener earth, but the lack of profitability for stakeholders meant that this was bound to happen. As I mentioned earlier, legal troubles seem to accompany cryptocurrencies. With the Merge, Ethereum is now even more secure, but it seems to have attracted some controversial attention.

Gagan Mahajan heads the Entertainment, Tech & Media Practice at Affine. Affine is a pioneer and a veteran in the data analytics industry and has worked with giants like Warner Bros Theatricals, Zee 5, Disney Studios, Sony, Epic, and many other marquee organizations. From game analytics and media and entertainment to travel & tourism, Affine has been instrumental in the success stories of many Fortune 500 global organizations; and is an expert in personalization science with its prowess in AI & ML.

Learn more about how Affine can revamp your business!

Film Wars- Stepping into the New Age of Content Viewership

It was 2009 when Avatar came out, and as a college student, I was proud to witness the spectacle of a technically-sound 3D movie, a cinematic experience of a lifetime.


More than 12 years later, Avatar still holds the record for the highest-grossing cinema of all time, with a worldwide collection of $2.847 Billion!

Avatar was a movie best watched in theatres for that astute and wholesome cinematic experience, an experience impossible to mimic even with the latest home entertainment options.

A decade later, Marvel positioned itself next in line as the second highest worldwide grosser of all time with Avengers Endgame in 2019.

A lot has changed in a decade. Marvel built its base for about a decade with its standalone comic-book movies and two other Avengers movies before achieving this feat at the box office. It was smart enough to understand the pulse of the audience and the growing interest in comic book-themed movies and TV shows.

Marvel became a giant over the years as the viewer preferences started to change, and a growing interest in live-action movies based on comic sources began to gain traction.

Cinephiles vs. Comic Book Fans

I am an avid cinephile, taking every chance to watch classics like The Godfather, Taxi Driver, Goodfellas, and Pulp Fiction, to name a few. So, when Martin Scorsese made an infamous statement comparing Marvel movies to amusement park rides, I could not help but agree with his view. A cookie-cutter template with abundant fan service and politically correct inclusive messaging is what phase 4 of Marvel looks like.


But personal opinion aside, the flavor of the season had changed, and comic book movies were minting money at the box office. This was an interesting phenomenon for someone in the Entertainment, Media, and Gaming industry. The gaming industry coined money and beat Hollywood revenue with Grand Theft Auto 5. During the pandemic, the gaming industry saw the highest footfalls among its peer industries. Unfortunately, that was not the case for theaters.

The Downfall of Theater Viewership


Theatrical revenue hit an all-time low after the shock of COVID, and the 2020 revenue numbers paint the picture perfectly.

But now that the dust has settled, the theater viewership has not. And one cannot blame the pandemic as the only reason anymore.

Consumer psychology has drastically changed post the COVID scenario, and according to a study, 55% of people preferred watching movies at home.

Theatrical vs. OTT Releases

Live-action movies that followed a pre-set template were safe for producers, worked well at the box office, and garnered enough business worldwide. Or at least, so I thought.

The pandemic came in like Thanos and, with a snap, reset the box office game.

OTTs had already become a vital part of everyone’s lives. I am subscribed to most OTTs, and it is a similar case amongst my peers and friends.

Due to the pandemic, most theaters were forced to shut down temporarily in 2020. With accruing interest in productions, studios had no other option but to release movies directly to OTTs. As people themselves could not freely wander out due to the lockdown norms, OTTs had become their go-to source of entertainment. During the lockdown, I caught up with Better Call Saul, Stranger Things, and many other TV shows and movies.

What did this do for the industry?

For one, the outdated exclusive window requirements saw a fall. Hollywood movies would only be released seven months after their theatrical release. This number was reduced to 45 days.

Studios are double dipping with theatrical and OTT releases, with the shorter window favoring OTT revenue a little more.

For theatrical releases, the first 38 days are crucial. But as I have stated earlier, Hollywood has struggled to provide a theatrical blockbuster for a long time outside of Marvel and DC.

High budgets, worldwide promotions, and a 50% split with exhibitors have added conundrums, whereas studios can have field-day OTT releases where they get to keep 80% of the revenue.

The Race to Beat Netflix

Netflix may have been the innovator in the streaming space, but as things stand now, I don’t see them having any particular edge over competitors compared to a decade ago.

Do not take my word for it; they are seeing a dip in subscribers and are even planning for a basic ad-included tier, which goes against what Netflix initially did. However, such is the current state of the saturated market amongst streaming platforms

Giants like Paramount, Disney, and HBO now have almost equal footing in the streaming space, and these players are veterans in the media and entertainment industry compared to Netflix. HBO Max will have an added company value of as much as $14 billion if it adds 11 million users at $144/year.

Jason Kilar, CEO of Warner Media, said the long-term value of HBO Max subscribers is worth more than the box office revenue of not just one of these movies but all of these movies!

While individual movies are devalued in their move to a streaming platform, the collection for the platform sees an increase in value, which is an ecosystem in itself and gives an incentive to the buyer to subscribe.

So, should Studios opt for streaming releases over Theaters?

As far as I see, there is no concrete answer here. No doubt theater releases have become tricky business-wise with so many parties involved, more rules, fewer shares, etc., but Top Gun Maverick just pulled off a sweet surprise with majestic global revenue, hitting a home run.

It is still doing great business, lasting over newer releases. Tom Cruise has consistently proved that he is a beast who can carry mega-blockbusters with ease, and the movie was a riot to watch in theatres and a treat after the extended lockdown. “Top Gun” is a fantastic movie without a superhero element, sans Tom Cruise!

Meanwhile, Marvel’s new phase is in a tricky spot. Thor: Love and Thunder got a lukewarm welcome, and She-Hulk, the TV show, looks dead on arrival. As I mentioned earlier, the message and preaching are getting old.

On the other hand, DC has canceled Batgirl and Supergirl projects worth over $70 million, which raises more questions. For one, the material may be questionable, which is what my research has led me to believe. Fans are not super impressed by political correctness and tend to pay attention to messages over the quality of content. People now are looking for pure entertainment in an overly politicized world.

The Right Time to Release Movies on OTT Post Theatrical Release

If a studio opts to prioritize OTT, it endures a significant financial squeeze in the short term to gain long-term financial security through scaled streaming.

This requires an astronomical investment and is a task in itself. Few services will survive the gauntlet. Those that do will be in a position to generate massive annual revenues.

Box office tickets and streaming revenues cannot be fairly compared as they are quite different.

Direct revenues from streaming may be lower, but they may be less volatile, replaced by more stable monthly subscriptions, like how boxed software suites have shifted to subscription services.

I think a portfolio-based approach is optimal for now and will bring balance on all fronts.

For value-based audiences and audiences for niche content, low production releases in large numbers are preferable. In such a scenario, immediate release to OTT is beneficial to studios, too, with more significant margins.

Warner Bros. Discovery has shifted the previous 45-day theatrical window put in place by former Warner Media CEO Jason Kilar, and now, new Warner Bros. films are no longer guaranteed to hit OTTs on that schedule.

The studio remains committed to the 45-day theatrical window but may extend the time between when a film is first released in theaters and when it is streaming on HBO Max.

My intuition tells me the release timing will vary based on the movie.

There is a dire need for studios to balance the revenue between Theaters and OTT platforms. With the audience split between Theater and OTT preferences, the right balance to capitalize on these potentials is the challenge of the hour.

The wave of change in viewer mentality has already hit, and the ecosystem of streaming platforms has made it even more challenging for studios. Still, double-dipping between theaters and Streaming platforms is possible if powered by an AI-powered data-backed solution.

Affine’s Movies/Shows cadence planning leverages box-office performance data with extensive reports and plans for a feasible and dynamic post-theatrical right-acting as a movie planning solution for OTT platforms. It proposes an ideal data-backed active window for OTT platforms to release a movie post-theatrical release and ensure maximum profits for the production house and significant benefits for SVOD/OTT platforms.

The payoff is that studios can double-dip their revenue while finding the ideal balance and catering to theatrical and streaming audiences.


I, for one, do not believe in a monopoly either on Streaming platforms or Theaters. Personal observations have led me to believe that the viewership behavior and content preferences have vastly changed over the years, but theaters and streaming can coexist peacefully.

Studios need to release their content backed by data to understand viewer preferences. Mega flicks with superior cinematic values are no brainer a theatrical release, but a conservatively produced movie might not share the same fate.

But this is just a general rule of thumb. Studios, with their multimillion-dollar projects, can’t just go based on this, a thoroughly researched strategy powered by AI catering to this viewer-driven market.

What does Affine bring to the table?

Affine is a pioneer and a veteran in the data analytics industry and has worked with giants like Warner Bros Theatricals, Zee 5, Disney Studios, Sony, Epic, and many other marquee organizations. From game analytics to media and entertainment, Affine has been instrumental in the success stories of many Fortune 500 global organizations; and is an expert in personalization science with its prowess in AI & ML.

Learn more about how Affine can revamp your film production business!

Copyright © 2024 Affine All Rights Reserved

Manas Agrawal

CEO & Co-Founder

Add Your Heading Text Here

Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s, when an unknown printer took a galley of type and scrambled it to make a type specimen book. It has survived not only five centuries, but also the leap into electronic typesetting, remaining essentially unchanged.