How Machine Learning is Boosting Businesses Industries such as healthcare, transportation, agriculture, cybersecurity, finance, transcription, marketing, retail, and education are the biggest benefactors of ML

Did you know that BFSI and fintech companies are now able to reduce NPAs, thanks to Machine learning-based data solutions? The usage of machine learning by a large global food manufacturer has resulted "in a 20% reduction in forecast errors, a 30% reduction in lost sales, a 30% reduction in product obsolescence and a 50% reduction in demand planners' workload." No doubt. Machine learning has emerged as a game-changer for industries. For the uninitiated, machine learning is a faction of Artificial Intelligence. It works towards training algorithms using data, enabling it to learn and adapt to perform specific tasks. In a jist, ML lets the AI system teach itself. It analyzes a large volume of datasets and identifies trends and patterns with ease as compared to manual analysis.

HELPFUL IN SEVERAL PROCESSES

This technology helps in several processes, with an important one of them being prediction. Prediction lies at the center of any business decision and it costs a lot. ML drastically reduces the cost of prediction. It automates routine tasks such as reporting, classification, auditing, and monitoring. A lot of unstructured data exists in businesses, be it large or small. Machine Learning helps the unstructured data get meaning and form. This structured data helps in the decision-making process such as in investments and strategies.

ML allows companies to create personalized customer engagement. For instance, it suggests your products or offerings based on your purchase history. Personalized recommendation often is a key for customers to return to the platform for future buys.

USAGE ACROSS VARIOUS INDUSTRIES

The useful technology finds it applications across industries spanning healthcare, transportation, agriculture, cybersecurity, finance, transcription, marketing, retail, and education. In the world of finance, Fintech start-up Digitap's AI/MLbased alternate data solutions are playing a crucial role in reducing Non-Performing Assets (NPAs). "By analyzing a wide range of data sources, including bank statements and social media, these solutions offer market-leading underwriting capabilities. This, in turn, minimizes risk and enhances the overall user life-cycle value, further illustrating how ML is revolutionizing customer experiences and personalizing services across industries," shares Nageen Kommu, Founder and CEO, Digitap.

"Think of (ML) as a financial advisor, tailoring solutions based out of individual needs using data, thus better helping our customers. It is capable of identifying patterns associated with fraud, security threats, and financial irregularities. This enhances risk management strategies, allows financial businesses to mitigate potential risks before they escalate. Overall, machine learning empowers businesses to move from reactive decision-making to proactive and strategic decision-making that is based on data, giving businesses a competitive edge in today's fastpaced and data-driven business landscape," adds Sunil Vashisth, CTO, LoanTap, a digital loan provider who is using Digitap's ML services.

Another fintech start-up Aye Finance, which focuses on micro-enterprises, is also leveraging ML. "We have successful ML models that support collection and bounce rate optimization, repeat loan eligibility screening and pricing, upfront credit profiling and underwriting, etc. An interesting point is that these models have withstood the Covid disruption and continued to adapt and perform well. For several of the deployments we are already at second or higher version of the models," said Tejamoy Ghosh, Head Data Science and Artificial Intelligence, Aye Finance.

Ghosh further adds that having deployed these models 4-5 years back, this system successfully withstood the pandemic and has "continued to adapt and perform well."

Other segments utilizing ML include stock market forecasting, fraud detection, risk management, credit scoring, and trade settlement process automation.

Many of your streaming platforms such as Netflix and Spotify who recommend shows, movies, videos, or songs do it by analyzing the songs you listen to and the watch history. ML algorithm identifies patterns and hence, suggests content you may enjoy. According to YouTube, the recommendations which a user gets are based on "over 80 billion pieces of information" about the user and they have been building this system since 2008.

Machine learning has directly impacted customer experiences or personalized services/products for Legalpay, a litigation financier. Deemed as India's only data-driven and tech-enabled alternativeinvestments platform, it specializes in legal and debt financing assets. "We use machine learning to find litigation financing deals by analyzing court documents and public records, assess the risk and return of each case it finances. By using historical data and predictive models, machine learning can help estimate the probability of winning, the expected ML empowers businesses to move from reactive decisionmaking to proactive and strategic decision-making that is based on data Sunil Vashisth, CTO, L oanTap duration, and the potential recovery amount of each case," shares Kundan Shahi, Founder, Legalpay.

Insurance firms are also actively leveraging machine learning in its processes. At large, they are using ML to provide custom products and premium to its customers and detect frauds in the claims raised. The algorithms also help in streamlining and automating the claims processing workflow, reducing manual interventions and making the process efficient.

In the manufacturing space, ML is helping in predict demand forecasts, quality assurance, predictive maintenance, and cost savings. VideoVerse, a visual AI start-up, through its flagship enterprise solution, Magnifi, analyze live sports games through ML and AI. "The advanced computer vision algorithms extract metadata, facilitating efficient library management, precise object tracking, and audio synchronization, further enhancing editing efficiency. Progressively, machine learning has remarkably reduced the TAT (turnaround time) to generate the highlights," shares Vinayak Shrivastav, Co-founder and CEO, VideoVerse. Magnifi also facilitates the smooth conversion of 16:9 videos into various aspect ratios while preserving the integrity of the focal points.

Chatbots are one of the major methods through which people have direct interactions with machine learning. Be it Amazon, Swiggy, or Kotak Mahindra Bank, all the big players offer chatbots to help customers solve their grievances.

A Harvard Business Review report stated that people preferred to solve problems themselves rather than talk to an agent. Legalpay's chatbot is powered by natural language processing (NLP) to provide ondemand assistance and guidance to its customers.

WHAT NEXT?

According to Statista, the market size for machine learning is expected to grow at a CAGR of 21.12 per cent, touching a market volume of USD 10.44 billion by 2030. However, it does have its share of consequences, namely biased data, high error susceptibility, and interpretation of results.

"We intend to use ML / AI to create more use cases for loan monitoring and for cross-sell and up-sell of products within the BFSI domain. The cross-sell / upsell use cases find applicability in domains outside BFSI as well and we are also exploring those," adds Kommu.

Shrivastav adds, "We are venturing into the domain of producing multilingual commentary audio through the utilisation of transcription and translation services. By diversifying our accessibility, we will create a more inclusive experience for the users. We are also working on implementing 'suggested highlight' recommendations and 'hype scores' using machine learning algorithms." AI, machine learning and data segments will also give a massive push to the Indian economy by creating 69 million new jobs by 2027 and ML is going to be a top role in the job creation as per the World Economic Forum. USD 10.44 bn The market size for Machine Learning is expected to grow at a CAGR of 21.12 per cent, touching a market volume of by 2030 Source: Statista