Ravindra Reddy Tamma
Data Scientist | ML Engineer | Deep Learning Engineer |
Techie | Programmer | Statistician | Cadet | Athlete
​
Phone:
+91 - 8500379936
+91 - 8074658323
​
Email:
​
Date of Birth:
March 03, 1995
​
Ravindra Reddy Tamma - Mr.Reliable!
​
Explore - Learn - Implement - Enhance - Contribute
​
Explore the available resources to solve the real world problem, Learn the basic principles and knowledge related to problem domain, Implement the solution to problem, Enhance the implemented solution to required saturation point, Contribute the new learnings to open source community.
​
EXPERIENCE
Data Scientist
PaySlip OCR
Amazon Textract
Survival Model
Risk Based Pricing
2022 Jul - Till Date
Senior Data Scientist
Jan 2020 - 2022 Jun
Collection Forecast
B2B Business
Medical Supplies
Holistic collection forecast solution using combination of Machine Learning models and statistical techniques to predict expected repayment amount from different obligor segments on a monthly basis
Prediction of BT
Home Loan
NBFC
BT (Balance Transfer) is a case where customer transfers the remaining loan balance from current portfolio to other bank. A Machine Learning model is build to score the active customers in the portfolio on quarterly basis to predict probability of BT in next 3 months.
Prediction of chances
of Pre-Mature birth
Health Care
A baby who is delivered earlier to 37th week (out of 40 weeks) is called pre-mature birth, A Machine Learning solution is developed to predict the probability of premature birth for a patient based on lab & scan parameters in the pregnancy journey
Auto ML Utility
DSO, Sales Forecast
Product - Backend
Built an Auto ML Utility that consumes invoice level data (sales & collections) and allows the users to process data and built customized tuned machine learning models.
Application Fraud
Generic Bureau Model
Equifax
Scorecard to predict the probability of fraud for a loan application based on bureau header data. As a generic bureau product the scorecard was tested against top 5 banks in india and model seems to capture 15-20% of frauds in top 2% population.
Fraud Scorecard
Consumer Loan
Private Bank
Scorecard to predict the probability of fraud for a loan application based on application data and bureau tradeline data. Fraud score card is aligned with internal risk scorecards to provide a wholistic strategy to efficiently implement the scorecard,
Analytics Consultant
Jan 2019 - Dec 2019
Deduping Customers
Customer Identity
Equifax
Built an algorithm to uniquely identify a customer based on multiple fields in the demographic information in the bureau. The algorithm involves in intelligent parsing of the indian addresses and assigning them a similarity score.
Income Estimation
Generic Bureau Product
Product - Backend
Model to estimate the income of a customer based on the active and closed loans reported to the bureau. The EMI is estimated based on outstanding balances and is used to model the income of a person.
System Engineer
Nov 2016 - Dec 2017
ETL Automation
Recommendation System
Private Bank
Implemented an Unix shell script to automatically extract data from RDBMS, transform it in to required format (creating features) and loading into hive database. The transformed data is fed into recommendation systems to get appropriate actionable for a customer
SKILLS
HOBBIES & INTERESTS
Bhagavadhgeetha
Volley Ball
Farming
EDUCATION
Post Graduate Diploma - Data Science
Jan 2018 - Dec 2018
Bachelor of Technology - Computer Science Jun 2012 - Jul 2016
Pre University - Maths, Physics, Chemistry Jun 2010 - Jul 2012
SSC / Matriculation
2010