Multi-Agentic RFP Proposal Platform
RFP documents in, grounded proposal drafts out — with expert review when requirements are ambiguous.
- Role
- AI Engineer
- Timeline
- Accellor
- Stack
- FastAPI, Service Bus, Multi-Agent AI, Azure
Problem
Responding to RFPs was slow and manual. Requirements had to be extracted, mapped to evidence, and drafted into proposals with compliance accuracy.
Solution
A multi-agent platform orchestrates extraction, grounding, and draft generation on FastAPI with Azure Service Bus jobs, plus a human-in-the-loop path for ambiguous requirements.
My Contribution
Architected the multi-agent flow, job orchestration, Hybrid RAG grounding across SharePoint/Salesforce/internal sources, and the SME review workflow that resumes automation after clarification.
Technology Stack
Results / Impact
HITL
expert review for ambiguous reqs
Grounded
evidence-backed compliance answers
Faster
proposal draft turnaround
Learnings
Full automation is the wrong goal for high-stakes proposals. Routing ambiguity to humans and resuming the agent graph is what made the system usable.