Vipin Mathew Email & Phone Number

Bengaluru, Karnataka, India

Part of Model Risk team

Bengaluru, Karnataka, India

• Part of Corporate Model Risk (CMoR) team as Senior Validator working with Retail and Commercial Credit Scoring, Fraud and Marketing model validation.• Responsible for validation of all Retail and Commercial Application and Behavioral Credit Risk Scorecards built in Wells Fargo. Working on validation of IFRS 9/CECL credit risk PD, LGD, EAD models.• Working as an Individual Contributor and leading a team of junior validators to perform Initial Validations, Revalidations, Targeted Scope Validations and Annual Reviews of Machine Learning models.• Responsible for providing support to the Head of Validation in communicating with regulators, auditors and senior management of the bank regarding model risk findings and their management.

Bengaluru, Karnataka, India

Worked as part of Fraud Data Science Team of Chase Consumer Banking, JP Morgan Chase.• Working on detecting whether an account was opened with the intention of fraud. Built Tree based models to interpret evolving fraud trends and segment customers and LSTM models to detect fraudulent accounts using the pattern of transactions in the account.• Increased by 3X, the end to end Machine learning model building pipeline performance in Spark using code optimizations for distributed computing environment.

Pune Area, India

Worked as part of Applied Machine Learning team of Barclaycard, the credit card division of Barclays. • Reduced average time spent in processing a trade document resulting in 6 FTE savings. The document processing engine developed using Python, Tesseract and deployed as a service using Docker, Flask and Apache HTTP server extracts Key-Value pairs from incoming documents and presents to the user for verification. Client-side OCR implemented using tesseract helps correct the errors with minimum effort.

Pune Area, India

• Development and application of machine learning algorithms to detect and prevent credit card frauds. Using pySpark for feature engineering on 1.5 Terabyte of credit card transactions and building fraud detection models using XGBoost and Deep Learning algorithms.• Working on improving Fraud Detection algorithms used by Barclays Partner Finance, one of the UK's largest providers of point-of-sale finance for retail companies and motor dealerships.• Implemented strategies for improving customer journeys through IVR system and reducing calls getting dropped to agents in Customer Contact Centers• Built a Chatbot on Facebook messenger platform which was released to selected customers as a pilot. Implemented Natural Language Processing models using NLTK, spaCy to understand a user's intent and automatically reply to queries related to Barclays

Pune

• Published in IEEE a paper titled “Sensor Health Monitoring using simple data driven approaches” which was also presented at Indian Control Conference 2016. The paper was cited as first reference in US Patent Number 10650616 which was filed by University of Connecticut and Toyota Motor Engineering.

Master Of Technology (Mtech), Digital Electronics And Advanced Communication

Bachelor Of Technology (B.Tech), Electronics And Communication Engineering

Education record