Vaibhav Choudhary

About Vaibhav Choudhary

IIT Bombay alumnus with 6+ years of experience in Data Science, GenAI, and BFSI Analytics. Currently at Publicis Sapient, I deliver production-scale GenAI solutions in banking and financial services.Scale: Engineered daily pipelines processing 15M+ transactions and 2M merchant names with zero incremental infrastructure cost.Accuracy and Cost Impact: Delivered 95%+ accuracy in merchant standardization at just $1 per 800 names.Compliance AI: Built RBI compliance chatbot (RAG + Gemma2:9B) and automated PII redaction pipelines securing medical and financial data with 85%+ accuracy.Client Growth: During a one-year consulting engagement, signed 3 new BFSI clients in 3 months, upsold into 3+ follow-on projects, and standardized proposal workflows.Leadership: Promoted within 6 months at AmEx, scaled AI/ML platform adoption 2.25× (4K → 9K users), mentored interns, and managed cross-functional delivery teams.Currently pursuing FRM Level 2 to deepen expertise in risk and financial analytics. Open to leadership-track roles in BFSI AI/ML, Fintech, and Consulting where I can drive measurable impact in compliance, risk, and client growth.Throughout my career till date, I have had a chance to:• Manage a team (8 people)• Work across clients and teams cross-functionally to build and manage unified enterprise capabilities• Build and automate end to end solutionsAchievements at my current job:• Delivered 95%+ accuracy in merchant name standardisation utilising paid LLM APIs with grounding,optimising cost at ~800 names for $1• Engineered and deployed daily processing logic for 15M+ transactions and 2M+ merchant names utilising existing infrastructure with 0 incremental costAchievements at my last job:• Successfully onboarded and extended client relationship with 4+ projects over 1+ year• Signed 3 new clients in a period of 3 months• Built a secure app interface to protect IP while improving usability for business users• Achieved a breakthrough in usage of Offline LLM moving away from paid API solutionsTools I have worked with:• BI Tools (Tableau, Tableau CRM, Kibana)• Cloud (GCP, Azure)• Python (ML, DL, NLP, LLM, OpenAI, SLM)• SQL (SQL Server, Oracle, MongoDB)• Big Data (Spark, Hive)

Experience

1. Manager - Applied Ai

   Publicis Sapient

   May 2025 — Present · Gurugram, IN

   Domain: BFSI || Tools used: Python, GCP, Vertex AI, SLM, LLM, Gen-AI, Business DevelopmentAchievements• Delivered 95%+ accuracy in merchant name standardization utilizing paid LLM APIs with grounding, optimizing cost at ~800 names for $1• Engineered and deployed daily processing logic for 15M+ transactions and 2M+ merchant names utilizing existing infrastructure with 0 incremental costProjects• Developed transformer-based models for spend and transaction categorization, attaining >90% accuracy (spend model covered 8 skewed categories)• Built a merchant name cleansing PoC by fine-tuning open-source SLM models, achieving 70%+ accuracyInitiatives • Established standardized ML-Ops practices, enabling streamlined, efficient deployment on cloud• Mentored and onboarded team members in GenAI pipeline operations and prompt engineering

Education

• St. Xavier's School, Jaipur

  10 + 2

  2001 — 2013

• Indian Institute of Technology, Bombay

  Minor, Computer Science

  2014 — 2019

• Indian Institute of Technology, Bombay

  Master's degree, Mechanical Engineering

  2014 — 2019

• Indian Institute of Technology, Bombay

  Bachelor's degree

  2014 — 2019