Research

The problems
behind the product.

Everything in the app rests on a few hard technical problems. This is what we work on, and where notes will be published.

  • Semantic document extraction

    In production

    Tender documents arrive as PDFs, spreadsheets, and scans whose structure plain text parsing destroys. We work on extraction that preserves meaning: tables, cell colors, requirement hierarchies, CV structure.

  • Capability-aware matching

    In production

    Keyword alerts drown teams in noise. We model what a firm can deliver, from its documents and win history, and score notices against that profile with reasons attached.

  • Procurement knowledge graphs

    Active work

    Buyers, suppliers, awards, and classifications form a graph that predicts who wins what. We build and query that graph across sources and classification systems such as CPV and DK021.

  • Agentic bid workflows

    Active work

    An agent that drafts, checks, and assembles bid material has to be verifiable. We study where autonomy helps and where the human must hold the pen.

Write-ups will appear here. Until then, progress ships on the changelog.