Summary
Overview
Work History
Education
Skills
Timeline
Generic

Siddharth Bisht

Manager - Data Science
Gurgaon

Summary

Data Science Manager with 5+ years of experience in credit risk, automation, and AI governance. Skilled in developing and governing large-scale ML models that evaluate exposures of ~$35B monthly, ensuring compliance with OCC/MRMG standards and enabling strategy teams through data-driven insights. Proficient in Python, Spark, SAS, SQL, XGBoost, SHAP and Azure AI. Recognized with Senior Vice President Awards and Star Award for enterprise-level contributions in model governance, risk frameworks, and responsible AI adoption.

Overview

6
6
years of professional experience

Work History

Manager - Data Science

American Express
Gurgaon
05.2022 - Current
  • Lead enterprise frameworks for variable selection (importance, stability, redundancy) across AI/ML models, including treatment and multi-target; research methods such as Boruta.
  • Direct segment-level tracking across approximately 200 models spanning credit, fraud, and marketing, defining business materiality, and governance dimensions.
  • Define vendor model governance using SHAP contribution and model PTI, expanding tracking to capture extended usage.
  • Support modelers on model risk findings, ensuring compliance with OCC and MRMG standards.
  • Led U.S. Consumer Risk Model development: data QC, vintage/variable selection, HPT, Interpretability, documentation, and deployment.
  • Directed simulation and back-scoring initiatives (2.1B+ records, 25MM+ customers, ~700M monthly transactions, evaluating exposures of ~$35B USD) to enable offline A/B testing and guide strategic decision-making across 10+ business teams.
  • Researched variable consolidation (credit line increase requests), improving parsimony without performance loss; analyzed 2x FRP enrollment growth (2022–2023) to assess model behavior.
  • Partnered with the CCO office on escalations, providing model explainability; consolidated the Adverse Action Framework, reducing approximately 150 to 40 rules, impacting millions of customers monthly.
  • Tracked liquidity risk model in 24 markets, monitoring dollar metrics and 60 DPB rates; collaborated with CCOs to expand segmentation (e.g., Italy CRIF, UK SBO, Mexico deferred payment).

Data Scientist

IBM
Bengaluru
10.2020 - 05.2022
  • Served as an SME for Azure Cognitive Services and Power Platform, delivering ML and cloud solutions to global clients.
  • Partnered with the IBM US AI Elite Team and a nonprofit to model debt vulnerability in low-income U.S. Communities using PLAID transaction data and credit risk models.
  • Built a hierarchical risk audit model (Python, DistilBERT, multinomial NB) to process over 100,000 documents, enabling 50% cost savings ($200,000–$300,000 annually).
  • Developed advanced ML solutions: Recommendation Engine (Autoencoders, OpenCV, ANN), Name Matching Model (LSH ensemble, ~0.03s/query), Email Processing Pipeline (OCR, Azure Form Recognizer, ~0.45s/doc), and Entity Extraction Model (NLP, Watson Discovery) for global clients.

Data Science Intern

IBM
Bengaluru
01.2020 - 07.2020
  • Built a synthetic data generation pipeline for invoice processing, creating semantic and syntactic datasets.
  • Developed a sequence matching and classification model (N-Gram, Laplacian Smoothing, LSTMs), reducing misclassifications by approximately 30%.
  • Integrated outputs into the document processing engine, lowering false negatives, and improving automation efficiency.

Education

Bachelor's of Technology - Computer Science

NIIT University
Neemrana
04.2001 -

Skills

Domain : Credit Risk Modelling, Liquidity Risk, Segment Tracking, Adverse Action Frameworks, Model Risk Management (OCC/MRMG), Vendor Model Governance, Responsible AI & Ethical AI

  • Machine Learning : Model Development & Validation, Variable Selection (Boruta, RFE), Explainability & Interpretability (PDP, ALE, SHAP), NLP (NLTK, SpaCy, DistilBERT), Deep Learning (TensorFlow, Keras, PyTorch), Scikit-learn, XGBoost

  • Programming & Data: Python, SAS, Pandas, NumPy, SciPy, PySpark, Hive, SQL, MongoDB, Redis

  • Visualization/Reporting: Matplotlib, Seaborn, Power BI

  • Cloud & DevOps: Microsoft Azure (ML, Serverless Functions, Cognitive Search, Cognitive Services, Web App, Cosmos DB, SQL Server), IBM Cloud (Watson ML Studio, Cloud Pak for Data, Watson Discovery, Knowledge Studio, AutoAI), Docker, Git

  • Deployment & Data: Model Scoring, Simulation Data, APIs, Cloud Platforms (Azure Cognitive Services, Power Platform)

Timeline

Manager - Data Science

American Express
05.2022 - Current

Data Scientist

IBM
10.2020 - 05.2022

Data Science Intern

IBM
01.2020 - 07.2020

Bachelor's of Technology - Computer Science

NIIT University
04.2001 -

Similar Profiles

CREATE PROFILE
Siddharth BishtManager - Data Science