AI has the potential to add $957 Bn, or 15 percent of India’s current gross value in 2035
In line with this opportunity, fintech startups have been seen adopting AI at a much wider scale
Many fintech startups are using AI to automate parts of operations such as data collection, document processing and customer onboarding
Jonathan Bill, cofounder and CEO of lending tech company CreditMate, believes that AI is not just the future but the present reality for the fintech industry. And that’s not the lone opinion that emphasises how much impact artificial intelligence will have on financial services, banking, credit and payments as more and more companies put AI technology to use for fintech.
According to the market research report published by P&S Intelligence, the global AI in retail market attained $720 Mn in 2018 and is predicted to witness a CAGR of 35.4% during the forecast period (2019–2024). The growth factors include increasing investments in AI by retail companies and expanding e-retail industry which is tapping into the efficiencies that automation has brought in to study consumer behaviour and capture relevant data through machine learning, natural language processing (NLP), and computer vision.
For instance, like Saket Anand, chief analytics officer, Lendingkart explains, at the core of any lending company’s business are credit models that work across several functions, including creating risk scores for customers, algorithms that decide what loan amounts they are eligible for, and personalised interest rates. AI is playing a crucial role in boosting the reach of lending tech companies.
Where Does AI Opportunity Lie In The Indian Fintech Market?
A recent study titled ‘State Of Enterprise AI In India 2019’, published by Analytics India Magazine in association with BRIDGEi2i, suggests that the Indian enterprise market for AI applications is estimated to be valued at $100 Mn, growing at 200-250% CAGR.
Another report by Accenture shows AI has the potential to add $957 Bn, or 15% of India’s current GDP to the economy by 2035. The combination of the technology, data and talent that make intelligent systems possible has reached critical mass, driving extraordinary growth in AI investment.
The Best Use Of AI In India’s Fintech Market?
With a number of fintech business models in place including the likes of neobanking and banking-as-a-service (BaaS), fintech startups have been been adopting AI algorithms and models at a much wider scale than ever before.
Here are 10 fintech startups using AI to innovate and streamline the daily processes and create algorithms in order to strengthen their data analytics and decision-making capabilities.
- Capital Float
Let’s dive into the details of each of these fintech startups, on how they are using AI in their operations:
Capital Float uses AI predominantly in risk assessment, collections, and marketing. Credit decisions such as which applicants should be given loans and when to offer a loan or a top-up amount and many others are determined by blending selective human insights with artificial intelligence.
The startup has designed intelligent, hybrid processes in collections that depend on AI and limited human involvement to preempt bad loans. AI models are applied in their marketing efforts to achieve improved customer targeting at a reduced cost of acquisition.
A significant number of functions in Walnut, a personal finance management app (acquired by Capital Float in 2018), involve AI engineering to ascertain the user’s expenditure behaviour and spending patterns.
Fraud detection, propensity models, risk analytics and virtual chat assistants are necessities of lending today. Much of the AI focus has been on onboarding customers. CreditMate is constantly innovating to mirror this development at the other end of the funnel in collections.
Collections is an intensely decision-driven process. Human agents are constantly deciding on whom to call or visit when and what to say. They resolve conflicts and figure out what triggers the user. Currently, this knowledge resides in the chaotic minds of a million agents as experience, but it’s being translated into AI systems as well.
Creditmate, using AI and machine learning, looks to extract all of this knowledge and make a single ultra-smart and evolving intelligence layer. Recommender systems match content to users. Behaviour models predict strategy for collection resolutions and Propensity Models drive efficiency. This helps in resolution of bad debts through automation, and significantly improves efficiency in last-mile or field collections. The goal is to use AI to reduce NPAs.
Coverfox uses AI algorithm-based insights to allow users to compare and choose from a range of plans across top insurance companies.
It is also integrating the process of policy issuance, endorsements, inspection and claims with insurance companies in order to make sales and post-sales service easier. Their team of data scientists analyses consumer interactions with various projects to provide meaningful trends.
The startup is working on cutting-edge technologies and algorithms to create a fully-automated lending platform over an API infrastructure. Its AI capabilities range from image recognition and OCR to automated onboarding, anomaly detection, identifying frauds and eliminating human bias.
This facilitates increased conversions and improved collections efficiency, deep learning to analyse speech, conversational intelligence in chatbots using NLP, and scorecards to extract value and reduce risk at every step of the process. Further, its Deep Learning and Technology Applications (DELTA) lab utilises alternative sources of data, deep learning networks, and advanced technologies to improve lending processes.
Some of its near-future research includes aim at the usage of AR-VR, drones, and facial expressions to deliver super high efficiency and understand user behaviour to determine risk.
Lendingkart’s proprietary algorithms take into account over 10K variables to evaluate the creditworthiness of customers. It processes about 7K qualified leads every day, resulting in an analysis of approximately 800 applications with 10K data points.
At Lendingkart, AI is being utilised for credit evaluation, quality lead scoring, and product interaction, among others. On the collections front, they are able to predict accounts that are likely to go delinquent in advance. This helps in forecasting manpower requirements for the collections team, as well as guide the team to follow up with seriously risky accounts before they turn delinquent.
Similarly, on the marketing front, they now know exactly which campaigns are bringing how much revenue and hence we are able to balance spend across channels, which is not possible if companies are using traditional marketing media.
mPokket aims to provide credit to underserved audiences that are excluded by the traditional financial system. In order to do so, it uses technology intensively for proprietary credit-scoring algorithms to gather and harness thousands of data points across demographics, social, behavioural, financial, and transactional factors to identify patterns.
In addition to the core underwriting process, it also uses AI/ML tools in various manifestations in multiple other processes – user acquisition, user onboarding, customer support, collections, and more.
Mswipe’s digital merchant onboarding solution called F2A2 is built in partnership with Mastercard. It’s Asia’s first such solution. It has been developed by Bengaluru-based enterprise tech startup Signzy and it is done through F2A2 App.
Field-force automation app or F2A2 enables Mswipe’s 4K-strong salesforce to digitally capture merchant profile information, KYC documents, and authenticate them in real-time against 40+ different government databases.
The paperless solution drastically reduces merchant onboarding time from three days to 30 minutes, thus enhancing Mswipe’s ability to meet regulatory and KYC guidelines while acquiring merchants and reduce onboarding risk. Not only it has reduced turnaround time for installation by 25%, but the solution is also helping Mswipe in enhancing overall sales efficiency and detect discrepancies and forgery cases.
A peer-to-peer (P2P) lending platform, PaisaDukan has been adopting AI and ML solutions for redefining lending operations. Post establishment of the basic features for the platform, PaisaDukan has an AI/ML-enabled digital lending product for rural areas to support local factors such as lower document availability with borrowers to promote rural lending for wider financial inclusion, women empowerment and last-mile credit facility.
Presently, AI/ML solutions embedded with P2P lending has helped PaisaDukan to address multiple operational challenges such as automation of credit assessment to better assist lenders on the platform about the creditworthiness of borrowers and increasing engagement on the platform. Algorithms are also helping its systems learn predictive responses, for improved customer experience in the future.
RenewBuy leverages open source technologies to build robust yet agile systems. Its infrastructure is backed by cloud computing and scalability is kept at the centre of all its technology decisions. Through a cost-efficient, customer-centric model, it has created a proprietary end-to-end digital insurance platform and digitised the entire insurance value chain.
AI tools help RenewBuy in data collection, figuring out the permutations and combinations for insurance premiums, and use of data analytics to help companies create smart underwriting rules. It also allows banks and other providers to co-create products with insurance companies backed by rich data and analytics.
The entire process of policy issuance and renewals is automated. It also uses AI for partner and customer servicing like endorsements and quote generation, co-creation of products, pre-sales, post-sales and customer service using bots.
The self-inspection tool within its app integrates with insurers’ underwriting logic engine and uses AI to approve or reject the case in real-time. For post-sales, non-financial implications underwriting are automated, which provides immediate solutions to customers with regards to policy servicing and claims processing.
Another lending startup Shubh Loans aims to democratise credit for millions of borrowers in India, who are not yet part of the formal credit system. Its main focus is to expand the availability of fair and transparent credit.
It is also building the next generation credit assessment, lending and risk management platform, using the latest technologies, including Big Data analytics and AI. Armed with real-time analytics and credit reports based on alternative data, it enables lenders to access unserved and underserved market segments.
The Future Of Fintech
According to Tushar Garimalla, chief growth officer at Capital Float, AI is rightly expected to play a major role in disrupting the future of fintech. “We will see a stabilisation of AI models and this will be the differentiator between competing fintech startups,” he told Inc42.
Indifi Technologies’s VP or analytics Gaurav Sharma told us that the success of fintech startups will depend on providing the best customer experience and the ability to correctly assess the creditworthiness of borrowers. According to him, the sophistication of algorithms used to turn data into insights and actions is going to be a decisive factor in winning the trust of customers for the fintech startups.
Whether it happens in the short or long term, one cannot deny the fact that AI will transform the Indian financial services industry in more ways than one. If used right, it can place Indian fintech startups on the global financial services map.