
Naman Jajani
@namanjajani
I build explainable AI and risk models that improve fraud detection, transparency, and model governance.
What I'm looking for
At Capital One, I engineer functional-decomposition explainability frameworks for black-box models, generating variance contributions, main-effect plots, and interaction heatmaps. I piloted the approach on fraud and AML XGBoost models, capturing over 97% and 99% variance.
I've validated commercial PD/LGD, CCAR, retail banking, GenAI, Agentic AI, and debit-card fraud models. My work includes identifying risks in a $120B PD monitoring portfolio, reviewing deployment code affecting a $3B portfolio, and developing a challenger fraud model that captured about $200K more fraud in one month.
I also build AI systems, including a LangGraph and DSPy multi-agent call-summarisation system and an Agentic Graph RAG football scouting pipeline. I work across machine learning, deep learning, NLP, statistical modelling, Python, SQL, Spark, and XGBoost.
Experience
Work history, roles, and key accomplishments
Engineered explainability frameworks for black-box models and led validation of fraud and AML models. Built multi-agentic systems for call summarisation and developed challenger models to improve fraud detection.
Validated PD and LGD models for commercial banking, enhancing methodologies for complex portfolios. Conducted statistical testing and flagged material risks in monitoring frameworks.
Developed a pipeline to estimate electricity theft using pattern recognition and statistical filtering. Optimised algorithm runtime from 23 seconds to 1.2 seconds per user.
Education
Degrees, certifications, and relevant coursework
Indian Institute of Technology Kharagpur
Bachelor of Technology, Chemical Engineering
2019 - 2023
Grade: 9.06/10
Pursued a Bachelor of Technology in Chemical Engineering with a Micro-Specialisation in AI, achieving a CGPA of 9.06/10.
Availability
Location
Authorized to work in
Skills
Interested in hiring Naman?
You can contact Naman and 90k+ other talented remote workers on Himalayas.
Message NamanGet matched with your dream remote job
Sign up now and join over 250,000+ remote workers who receive personalized job alerts, curated job matches, and more for free!
