Turning data into production AI, NLP, MLOps, and API-first systems for regulated use cases in FinTech and beyond.
- Data Scientist and Software Developer focused on end-to-end ML delivery: from data pipelines to APIs and monitoring in production.
- Currently exploring LLMs and retrieval-augmented systems, and integrating them into scalable backends.
- Passionate about clean interfaces between data science and engineering: FastAPI, Azure Functions, CI/CD, and model health.
- Automated KYC/ID verification, cutting manual review time by 85% via Azure Document Intelligence + custom NLP.
- Reduced model degradation incidents by 40% with drift and data-quality monitoring (Evidently + CI/CD).
- Productionized BERT-based classification on Azure (containerized FastAPI), lifting accuracy by 8% and lowering latency by 30%.
- KYC/Onboarding Automation — Azure Document Intelligence + custom NLP to accelerate lender onboarding and cut manual checks.
- RAG FAQ Chatbot — Streamlit + LangChain + Azure AI for exec-grade answers, reducing escalations by 70%.
- BERT API on Azure — Containerized FastAPI service with latency and accuracy improvements in production.
- Income Verification Bot — Azure Functions + React integration to reduce manual verification workload.
- NLP Machine learning model - Developed and deployed ML classification model in Azure ML Studio, improving Underwriter efficiency.
- LLM orchestration patterns for enterprise workflows
- Robust evaluation: drift, data quality, and CI for ML
- Document intelligence for financial data extraction
- Email: gavhaneyuvraj54@gmail.com
