Deal flow, underwriting, and investor-ready sharing
A portal to manage listings, comps, docs, and pipeline updates with permissioned access.
Overview
Findings
Local RAG lab indexes dealroom data and uploads for cited retrieval.
Search and chat surface grounding metrics: score, source diversity, citation coverage.
Answers flag unsupported claims to keep responses audit-ready.
Notes
Runs offline with a JSON vector store; OpenAI is optional for richer answers.
Uploads support txt, md/markdown, and json files.
Re-indexing is safe and replaces stale chunks.
What it does
A standalone Dealroom demo with a local RAG lab that indexes dataset briefs and uploads, then answers with citations and evaluation checks.
Screenshots
Workflow
Step 01
IndexingSeed the dataset pack or upload docs (txt, md, json) with metadata.
Step 02
IndexingClean text and slice into 350-word segments for retrieval.
Step 03
IndexingGenerate embeddings and store vectors in a local JSON index.
Step 04
QueryTop-k semantic search returns the most relevant chunks.
Step 05
QueryLLM drafts the response and attaches citations for each claim.
Step 06
QueryGrounding score, citation coverage, and unsupported claim flags.
What we built
Value delivered
Outcomes
Decision speed
Fewer bottlenecks with a shared deal view
Repeatability
Standard underwriting snapshots
Sharing
Permissioned investor portal
Docs
Versioned vault + secure links
Demo
Get a tailored walkthrough for your team.
