The problem Algogator solves

In today’s digital world, a world that open handedly is embracing financial inclusiveness, thin file customers and NTC customers are still finding it hard to feel connected with the system of lenders or LSPs. The LSPs are forced to look in the other direction, to at least a minor chunk of the population, owing to the lack of capacity to identify credible, credit worthy customers.
We Thinkers @Think360 wants to clear the waters around this lack of trustworthiness that is brought into picture by using technology advancements in the field of Artificial Intelligence and Machine Learning. As account aggregators are becoming a reality that can totally wipe out the lack of credibility of information, the next big question the LSPs and Lenders are going to face is, how this information can be broken down to gauge the credit worthiness of each individual.
Todays Online Lending Business in India is becoming bigger and growing day by day and requests for loans and financial support by both individuals and businesses are growing at a staggering 24% globally. The Velocity of data is expected to keep growing in India as well.
Nowadays, Online lending businesses rely on alternative sources of data like SMS and Emails to gauge the trustworthiness of a customer. Only a very few TSPs in the market has so far got the math right in terms of cracking the ability to gauge customers from their banking statements when the velocity and volume is stunningly huge. With our understanding of the technology and the experience that we have gathered around in the past years working in this exact field, this is the problem that we seek to solve, to remove that one hurdle of inability to gauge creditworthiness of individuals accurately from banking statements when velocity and volume are high from the system. To prepare the road for the business that seek to onboard the account aggregator ecosystem, making lives all around us easier

Challenges we ran into

1. Amalgamation of data: The major challenge the team had faced while prepping for the hackathon was the extraction of data from the SETU Sandbox Environment and relating it with the PhonePe pulse data. The SETU sandbox data was extracted after successfully connecting with the APIs that are responsible for exchanging this data.
   The PhonePe pulse data was downloaded directly. Some effort had to be put into understanding both these data. The major issues here were in aggregating and converting the SETU Sandbox data into understandable data points. Post this, there was trouble in connecting this broken-down data with the PhonePe Pulse data. PhonePe Pulse Data was demographic data, only after identifying the individual’s data pertaining to their addresses, we could come out with a connection point.

2. Timeframe of the Data: After securely downloading the SETU sandbox data, on analysis, it was found out that all data points that was downloaded were pertaining to one individual customer and the entire data that was downloaded was only for only two years. This we believe kind of affected our understanding on the data exchange part of the ecosystem. The capacity of the system to ping multiple user data and the scenario could not be understood.

3. Data Insufficiency: The data from SETU sandbox had only a few insights to generate especially because of the lack of financial data from domains like insurance. Mutual Funds, Credit Cards. This significantly reduced our capability to gauge the entire financial profile of the customer. A major negative incident in financial background of customers is when an individual miss’s payment of an EMI/Credit card due. Unable to compute with accuracy theses data points might have affected outputs and insights.