Tathagata Gupta

About Tathagata Gupta

Leading the initiative of developing AI/ML challenger models across Financial Crimes, Fraud, Operations and Technology models. Extensive experience in developing CNN, RNN, GAN, NLP, LSTM, Naive Bayes, Gradient Boosting algos in Python using packages Tensorflow, Keras, Pyspark. Implementing Apache Spark algos for models involving huge data. Leading validation on ensemble models such as Isolation forest and K-Means and on a separate instance involving ensemble model based on Random Forest and Gradient Boosting Machine (GBM).Developed benchmark model using GAN (Generative Adversarial Network) for Operational Risk AMA model. Validated fraud model based on stacked ensemble methodology (Random Forest and Xgboost). Developed benchmark (AI/ML) model for operational risk champion model using deep learning. Developed a Natural Language Processing (NLP) model using tensorflow/Keras in Python. Used NLTK package and developed word embeddings using a Word2Vec model with skip gram approach and Noise Contrastive Estimation (NCE) followed by a convolution neural network (CNN). Validation of market risk models for BAU as well CCAR that includes models like IRC, IDL, CRM etc. CCAR GMS scenario model validation. CVA Loss model for CCAR. Scenario Computation model for different asset classes using full revaluation, partial revaluation and sensitivity based approach.PPNR model validation for CCAR. Validated Individual Risk Rating (Scorecard) model having a quantitative logistic part from PCA(Principal Component Analysis) and a qualitative part based on Non linear mixed model in SAS. Helped in validation of economic forecasting model in strategic analytics for HELOC, HELOAN, Credit Cards, Dealer Financial Services. Working on validation of OTTI (Other Than Temporary Impairment) from CMBS portfolio in Trepp Analytics.Quantitative Modeling using SAS for CCAR( Comprehensive Capital Analysis and Review ) process.PPNR( Pre Provision Net-Revenue) items - Non Interest Income and expense heads Modeling and Forecasting.

Experience

1. Executive Director - Quant Modeling Director (Gen Ai and Ml Models)

   JPMorganChase

   Nov 2023 — Present · NY, US

Education

• Devi Ahilya Vishwavidyalaya

  MBA, Major-Finance, Minor - Marketing

  2009 — 2011

• UNC Charlotte Belk College of Business

  MS(Mathematical Finance), Derivatives pricing, Risk Management

  2012 — 2013

Skills

• R Language
• Garch Modeling
• Financial Modeling
• Monte Carlo Simulation
• Finance
• Sql
• Quantitative Finance
• Corporate Finance
• R
• Fixed Income
• Microsoft Office
• Hedge Funds
• Derivatives
• Portfolio Management
• Equities
• Spss
• C++
• Investments
• Sas
• Ms-Office
• Financial Risk
• Asset Management