Jay Patel

About Jay Patel

I\'m a data scientist with 6+ years working at the intersection of ML, data engineering, and financial risk by building systems that help banks and fintechs detect fraud, understand credit risk, and make faster, better-informed decisions.At EXL Services, I lead fraud analytics for a digital banking client processing 50M+ monthly transactions ($2B volume). My work spans the full data stack: from building Airflow + dbt + Snowflake pipelines that eliminated 90% of manual reporting, to applying statistical segmentation that recalibrated fraud detection thresholds and reduced losses by approximately $600K/month. I don\'t just surface insights but build the infrastructure to make them reliable and repeatable.Before EXL, at Better Mortgage, I analyzed 100K+ mortgage applications to uncover trends in borrower risk, approval, and prepayment behavior insights that helped refine risk models and improve portfolio performance.On the ML side, I\'ve built end-to-end pipelines from scratch. PropNavigator, my real estate intelligence platform that takes raw property listings through Selenium scraping, feature engineering, ensemble model training (XGBoost, Random Forest, SVR), and live Streamlit deployment. The price prediction model achieved 12.98% MAPE on records. It\'s live.My focus has been on helping product and risk teams to make faster, data-driven decisions, whether it’s stopping bad actors early, improving rule precision, or streamlining how fraud cases are managed operationally.My background is a bit unconventional: I started as a mechanical engineer, earned an MBA in Business Analytics, and spent my career steadily moving deeper into data science from Tableau dashboards, to SQL analytics, to production ML pipelines.What I\'m looking for: a core data scientist role at a tech-forward company where I can work on technically challenging problems, grow alongside strong ML practitioners, and build models that actually ship. Portfolio: datascienceportfol.io/pateljay Open to: Data Scientist | Senior Data Analyst | ML Engineer roles

Experience

1. Lead Assistant Manager

   EXL

   Jan 2025 — Present

   Leading fraud analytics and ML for a top US digital bank — a platform processing 50M+ monthly transactions ($2B volume).Key contributions:• Reduced fraud losses by ~$600K/month through Mastercard score band segmentation and threshold recalibration (fraud BPS: 7.6 → 5.2)• Eliminated 90% of manual reporting by automating a production Airflow pipeline (API → AWS S3 → Snowflake) with incremental dbt models• Cut first-year credit card loss exposure by 10–12% through a post-approval performance framework tracking delinquency and charge-off metrics• Built unified Snowflake analytical views connecting authorization, fraud loss, and dispute data across three datamarts

Education

• Institute of Management Technology, Nagpur

  Master of Business Administration - MBA, Business Analytics and Finance

  2018 — 2020

• A D Patel Institute of Technology, CVM University

  BE - Bachelor of Engineering, Mechanical Engineering

  2011 — 2015

• School of Achievers

  Science, Higher Secondary

  2010 — 2011