AgileRadar: AI-Assisted Public Tender Scouting & Proposals
An internal platform that scans Chile’s public procurement portal, scores each tender against our capabilities and drafts the first proposal.
01. The Challenge
Scouting public tenders means reading 50 to 100 pages of PDF or Word documentation per bid. Assessing technical fit by hand takes days, and high-value opportunities get missed.
02. The Solution
We built a pre-qualification workflow on top of Mercado Público: a real-time Kanban board over WebSockets, AI fit-scoring from 0 to 100 with a written justification, a drag-and-drop consolidator for PDF and Word attachments, and a draft engine for technical and commercial proposals. Three cascading AI providers (Groq, Gemini and local Ollama) keep the pipeline running if one fails.
03. The Result
An internal tool in daily use that turns days of manual document reading into an assisted review, so the team decides which tenders to pursue and starts from a draft instead of a blank page.
