Lead Assistant Manager (Data Science - Generative AI Engineer) at EXL (2025-11 – Present)
- Developed Generative AI and agent-based workflows using LangChain, LangGraph, AutoGPT, and ReAct to automate agent creation from transcripts and enable real-time decision support within enterprise systems.
- Built LLM applications including RAG pipelines with vector databases for text summarization, content generation, and data enrichment across insurance, healthcare, and finance domains.
- Fine-tuned transformer models and developed NLP pipelines for information extraction, summarization, sentiment analysis, intent classification, and entity extraction.
- Supported cloud AI deployments on AWS, Azure, and GCP, handling data pipelines, model execution, monitoring, and evaluation while adhering to responsible AI guidelines.
- Performed data preprocessing and feature engineering using Python and SQL
- Collaborated cross-functionally with data scientists, engineers, and business teams.
Data Scientist at Reliance Industries (RIL) (2023-06 – 2025-11)
Leading the development and production deployment of advanced analytics and Generative AI solutions, delivering measurable business impact through predictive modeling, demand forecasting, operational automation, and intelligent assistants. Spearheaded 6 high-impact Data Science and Generative AI projects from concept production.
- Demand Forecasting and Transit Time Prediction models (time-series & ML) for supply chain optimization
- OMS Assistant: Multi-agent RAG system using LangGraph, CrewAI, Nomic embeddings (Ollama), ChromaDB, FastAPI backend on Azure VM with WebSockets; reduced support tickets by 30%
- WellBot HR: RAG-based employee self-service AI for policy & well-being queries, significantly reducing manual HR interventions
- SelfGPT: Personal RAG assistant using LangChain, OpenAI, and ChromaDB for context-aware responses from private documents
- AI-Powered Interview Practice Assistant: End-to-end agent that ingests resumes, generates questions, evaluates responses, and provides feedback
- Designed and productionized ML/LLM workflows using Docker (GPU/CUDA), Azure Container Registry, FastAPI, Streamlit, and monitoring best practices; established reusable MLOps templates adopted across teams.
- Built and delivered 50+ enterprise Tableau dashboards (desktop and mobile versions) consumed by operations, supply chain, and leadership teams for real-time decision making.
- Developed Python automation suite for email-based report ingestion, processing, and distribution - saving stakeholders 3+ hours daily.
- Served as internal SME for Generative AI, RAG architectures, LangChain/LangGraph, CrewAI, prompt engineering, and production deployment of LLM applications; conducted multiple knowledge-sharing sessions.
- Collaborated closely with business units and digital teams to translate complex analytical and AI outputs into actionable insights and user-friendly applications.
Associate Software Engineer at Accenture (2021-10 – 2022-08)
Worked on pre-migration and post-migration tasks in cloud platforms like google cloud engine.
Trainee Engineer at Cartesian Consulting (2021-07 – 2021-09)
Worked on maintenance and support of client's databases using technologies like MySQL and Linux
