What it does
In one paragraph.
Ingests RFPs, BRDs, TRDs, and SoWs in PDF/DOCX, classifies sections, extracts structured data and routes downstream to the rest of the suite.
< 90s
to parse a 400-page RFP
4
core algorithms
How it works
The pipeline.
PDF.js / Mammoth.js parse → layout-aware tokenisation → section classifier (TF-IDF + heuristics) → entity extraction (regex + NER) → structured JSON.
Core algorithms
- Layout segmentation
- TF-IDF section classifier
- Regex entity miner
- Table reconstruction
Inputs
- PDF / DOCX / TXT
- Free-text paste
Outputs
- Sectioned document map
- Entities (modules, T-codes, tables, dates)
- Routing manifest for Tools 02–14
Get started
See this tool on your own data.
Let's map your highest-leverage AI opportunities in a 30-minute strategy call.