The Dvara Group (formerly IFMR Trust) has entered into a Memorandum of Understanding with The Robert Bosch Centre for Data Science and Artificial Intelligence (RBC-DSAI) at IIT Madras. The collaboration would see the Dvara group entities - Dvara KGFS (a rural wealth management institution), Dvara Trust, and Dvara Research - jointly work with IIT Madras on initiatives that aim to advance financial access to low-income households.

Under the partnership, the RBC-DSAI would work with Dvara KGFS to deploy state of the art statistical techniques and advanced analytical tools to create insights and decision support systems that would aid and benefit over 8 lakh Dvara KGFS customers in remote rural India.

Samir Shah, Executive Vice Chair and Group President of Dvara Trust, said, “With this partnership, we aim to ensure a more systematic approach in managing wealth for low-income households which include farmers, rural entrepreneurs and vulnerable families who cannot afford to make bad financial decisions.”

Prof Nandan Sudarsanam, Department of Management Studies, who is the lead investigator from IIT Madras, said, “Business challenges and data availability in the underbanked setting can be significantly different from standard retail banking. As a result, we find that off-the-shelf models and analytics techniques which are typical in traditional banking, are ill-equipped to address the needs of an institution working in the underbanked space. By teaming up with Dvara to develop data-driven analytics and business intelligence in such a context, we aim to address an important lacuna which can have significant societal impact.”

The agreement envisages jointly working towards the creation of a prediction engine that can infer crop type and yield based on satellite images. Such an engine will significantly improve the way in which risk is assessed before lending and will be of tremendous value in determining the type of risks and assets, a release added. The collaboration will see the exploration of research questions around borrower stress, and informal lending through analysis of behavioural and demographic patterns deciphered through large datasets.