GenAI & Agentic AI Consultant · Full-Stack Builder
Agentic RAG, multi-agent orchestration and full-stack AI. Designed, built and shipped end to end.
I take GenAI engagements on my own, from scoping through build, deployment and handover. Currently also building Cardoo, a digital identity app for India 🇮🇳
| Capability | What it means |
|---|---|
| Agentic RAG | Hybrid retrieval, cross-encoder reranking and metadata filtering over your own knowledge base |
| Multi-Agent Orchestration | Stateful, tool-calling agents that plan, act and hold context across long-running workflows |
| LLM Application Backends | Typed APIs, structured outputs, streaming and schema validation wrapped around the model layer |
| Document Intelligence | OCR and layout-aware parsing that turns unstructured documents into clean, queryable data |
Every system ships with cost controls, schema validation and a human in the loop, because a demo that impresses is not the same as one you can run on Monday.
| Engagement | What it does |
|---|---|
| Legal Agentic RAG |
Production RAG over 50,000+ contract and legal PDFs. Hybrid BM25 plus dense retrieval with cross-encoder reranking, which cut the hallucination rate measurably. |
| AI Hiring Automation |
A 7-stage pipeline from job post to onboarding that removed roughly 95% of the HR workload per hire, leaving just 3 human decision points. |
| RFQ to Quotation |
Email in, techno-commercial offer out. Nothing reaches a customer without human approval. |
Techniques: hybrid BM25 and dense retrieval · cross-encoder reranking · metadata filtering · tool-calling agents · structured outputs · schema validation · OCR and layout-aware parsing · tracing, evaluation and token budgets
Techniques: hybrid BM25 and dense retrieval · cross-encoder reranking · metadata filtering · tool-calling agents · structured outputs · schema validation · OCR and layout-aware parsing · tracing, evaluation and token budgets
"AI isn't about generating text. It's about systems that reason, act, and hold up in production."