Submit a talk on data

Submit talks on data engineering, data science, machine learning, big data and analytics through the year – 2019

Crisptalk - Rule-based SMS analysis for Credit Risk Insights

Submitted by Ashish Mukherjee (@ashish-m-zl) on Tuesday, 12 June 2018


Technical level




Vote on this proposal

Login to vote

Total votes:  +1


The lending space is rapidly growing in the fintech arena in India. The challenge faced by this sector is operational scaling of the business to an unprecedented level without adding large amount of manpower. In order to meet this challenge, the traditional process must be replaced in part or whole by digital data sources, one of which is SMS. The solution designed is able to analyse SMS data and find out critical data points which can provide key business insight. The rule-based system designed is highly flexible in being adapted to variations in the data and currently performs at approximately 85% accuracy. The solution is currently in operation as part of ZipLoan’s underwriting process. The talk covers the entire journey from developing an understanding of the problem to the eventual solution currently in production and ongoing work towards future improvement. Business understanding is blended along with necessary technical details in expounding the final shape of the solution.


  • Introduction to business need and problem space
  • Use Cases addressed by the solution
  • Technology options and challenges
  • How we solved the problem
  • Pros and Cons of the solution
  • Onging work



Speaker bio

Ashish Mukherjee is V.P at ZipLoan Labs, which is a fintech startup in the lending space. He has more than 18 years experience building systems at scale across domains and companies. He has worked for technology giants like Yahoo and Monster and carries rich startup experience as well. Further, he has a defensive publication to his name in the e-recruitment domain. His experience spans application development, devops, Big Data and NLP. At present, he is applying his skills in a team focused on data science at ZipLoan.



Login with Twitter or Google to leave a comment