Ask your contract a question. Get the answer in seconds.
AI Document Intelligence is the flagship AI capability of ProjectSathi — a Retrieval-Augmented-Generation (RAG) engine built on top of a graph index of your project documents. Every contract, drawing, letter, MoM, and BOQ gets parsed, chunked, embedded, and indexed. Then your team asks plain-English questions and gets answers with source citations from the original PDFs. Built for Enterprise tier; flagged on every project.
What’s inside
5 built-in sub-features
Every capability ships with the module. Nothing charged separately.
Dataset Overview
Live KPIs — total files, indexed chunks, vector DB size, active pipelines, parse success rate.
Files
Browse and manage every ingested document — re-index, mark sensitive, version-track.
Retrieval Testing
Test prompts against the corpus before rolling out to users. Tune chunking, top-K, rerankers.
Logs
Pipeline logs for every ingestion run — parse success/failure, latency, embedding model used.
Configuration
Configure embedding models, chunking strategies, rerankers, and retrieval policies.
How it works
From first login to running in days, not months.
Upload your contract library
Drag in PDFs — contracts, drawings, addenda, letters. RAG-Graph indexes every page.
Ask in plain English
Type any question. The AI retrieves the exact clause, page number, and source document.
Audit the answer
Every response cites source documents. Your team checks the original, not a summary.
Available on
Included in your annual contract — no module add-on fees.
Related modules
These modules work alongside AI Document Intelligence in a single deployment.
See AI Document Intelligence on your sites.
We’ll walk through this module with data shaped exactly like yours — manpower, sites, BOQ structure.