As India slowly emerges from the second wave of COVID19, the after effects will linger on for sometime. Aside from taking a brutal beating physically, several Indians have been left struggling to cope financially. Lockdowns have adversely impacted coffers, and rebuilding depleted savings will take time. India has nearly 600 million people who earn less than Rs 300 a day and lack access to structured financial services. 

This is when having a goal-oriented approach to savings will really help. Puneet Gupta and Sucharita Mukherjee, former employees of IFMR, started Kaleidofin in 2017 to assist underbanked individuals with goal-based financial solutions. A career finance professional, Gupta spent time in the microfinance division of ICICI Bank where his role was centered around strengthening delivery of financial services to low-income households. Following a six year stint at ICICI, Gupta spent nearly a decade at IFMR where he focused on building models that could maximise the financial wellbeing of low-income households, with a keen emphasis on developing retail and technology offerings. 

Following demonetisation in 2016, the financial landscape in India underwent a drastic change. Gupta noted that while underbanked individuals were being given access to Jan Dhan accounts, e-wallets and Aadhar-based financial payment options, an inclusive and cohesive mechanism was missing. Along with Mukherjee, who was his colleague at IFMR, Gupta started Kaleidofin that would leverage emerging technologies like AI and ML to bring together the dynamic elements of digital financial services, and deliver an inclusive banking experience for all. 

"We knew we had to rely heavily on AI and ML to create a suite of offerings that would deliver effective methods of financial savings. One of the key differentiators for us is to communicate with our customers in a language they understand - most financial service providers would bandy terms like short term bonds and savings schemes, but not many truly understood why so many customers struggled with saving money. We eventually realised that the way we conveyed the significance of saving money was as important as the solutions we could offer," explains Gupta. 

Designed to suit the needs of the underbanked, each Kaleidofin Goal solution is a combination of savings, investment, insurance and loans. This helps the underbanked plan and save towards goals more intelligently. A partner network of 36 entities comprising microfinance institutions, NGOs and corporates provide the support to reach customers. A customer can choose an amount as low as 500 per month to save. In addition, there is no penalty if the customer cannot save for a few months. Every customer profile is meticulously etched out, with data on his/her financial goals, sources of financial vulnerability, risk appetite and their overarching goals, in addition to demographic profile, source of income, asset ownership. Kaleidofin also leverages the full India stack to provide customer and product partners with seamless integration to the platform, including e Know Your Customer (eKYC), eSignature (eSign) and Electronic National Automated Clearing House (eNACH). 

All this data leads to creating a digital 'persona' of every customer on the Kaleidofin platform, for whom tailored financial savings plans can be recommended, with the help of algorithms and machine learning. Popular savings plans are Lakshya, Ummeed and Udaan - all suited to support a range of personas ranging from savings-averse and volatile to meticulous and risk-friendly. Other areas where AI is used include document detection using OCR, automated Aadhar redaction, automated anomaly detection on complex NACH forms via handwriting recognition and facial recognition to improve KYC compliance and accuracy. 

During the pandemic, Kaleidofin also introduced a new service called KiScore, a data driven risk scoring model for customers to help MFIs make better credit decisions. KiScore relies on quantitative characteristics of the borrower’s loan repayment history. The outcome of the KiScore credit risk model is a continuous default probability distribution. Each customer is evaluated based on several raw information(variables) and mined information (derived variables) from data to arrive at a score that takes values between 0 and 1. KiScore relies on quantitative characteristics of the borrower’s loan repayment history. The output from a KiScore risk scoring model typically describes the default behaviour of a customer i.e., how likely is it for the customer to pay his/her loan back. KiScoreTM had been used to score over 750,000+ customers.

As of February 2021, Kaleidofin Goals had over 100,000 customers, spread across 10 states such as Bihar, Jharkhand, TamilNadu, Gujarat, Rajasthan and Madhya Pradesh.

As of May 2020, 75% of customers continued to deposit their monthly contributions through Kaleidofin's digital payment systems. The startup aims reach a cumulative goal amount of $59 million for its customer network. Kaleidofin has raised $5.1 million as Series A funding from Flourish, Oiko Credit, Omidyar Fund, Blume Ventures, Bharat Fund.

Gupta's plans in the coming months include developing an app and chatbot supported by regional languages to penetrate the last mile more effectively, and continue building solutions leveraging intelligence built on customer insights.