Tyche Institute · Procurement Signals

Where public procurement reveals AI, PQC, and PKI.

RHR parses public procurement notices from open eForms/UBL data — not UI scraping — and packages AI, post-quantum-cryptography, and public-key-infrastructure signal hits as receipted evidence objects, so a public notice can become an inspectable signal with provenance intact.

Baseline signal identities91distinct AI/PQC/PKI signals tracked across notices
Topic classes3AI · post-quantum cryptography · public-key infrastructure
SourceOpen dataeForms / UBL notice exports, not UI scraping
Evidence packagingSigned receiptsbyte- and hash-verified snapshots
How it works

From public notice to receipted signal

Parse open data

Normalizes eForms notices, award notices, and procurement plans from public XML exports — well-documented through TED eForms/UBL — rather than scraping a UI.

Classify signals

A config-driven classifier emits AI, PQC, and PKI topic hits with snippets. Broad PKI certificate terms are noisy and are routed to manual review.

Receipt the snapshot

Each tracked signal is packaged with a signed-snapshot receipt that verifies the package bytes and declared hashes — provenance, not legal status.

Scope

RHR reports public procurement signals with verifiable provenance. Receipts verify package bytes and declared hashes only; they do not turn an observation into a legal, compliance, or supervisory finding, do not guarantee endpoint stability, and do not imply that open data includes every tender attachment. Broad certificate terms require manual review. Counts are attributed to this observatory and never fused with other surfaces.

© Tyche Institute · Tallinn, Estonia · Part of the AI trust-stack evidence ecosystem. · Flagship article