# ABoringBot

> ABoringBot builds "boring" bots. Each bot turns one source of messy public data into clean, dated, verifiable records, delivered to people (weekly email) and to AI agents (described in /llms.txt). First bot live: EU public-tender awards from TED: who wins, at what price, with how many bids.

Data is derived from TED (Tenders Electronic Daily), (c) European Union, reused under Commission Decision 2011/833/EU. Data modified from the original: normalised, deduplicated, currency-converted where stated. Not affiliated with the EU Publications Office. Every record links to its TED notice and carries an evidence level ("complete" or "partial").

Current dataset (as of 20 Oct 2026): 22,508 award rows from 4,631 notices published 2026-08-21 to 2026-10-20, countries ESP, DEU, ITA. 1,210 rows (5.4%) are complete (about 1 in 19); the rest are partial. Framework agreements (16,992 rows) are marked and excluded from discounts and from complete evidence. 5 rows with award-to-publication delay over 1,000 days are flagged.

Record fields: notice_id, row, rows_in_notice, country, cpv_division, buyer, winner, awarded_value_eur, estimated_value_eur, discount_pct, tenders_received, award_date, publication_date, evidence, framework_agreement, retrieved_at, url.

## Pages

- [Homepage](https://aboringbot.com/): what the bots do, sample records, evidence policy

## Data access

- Weekly email report: free sample, sign up on the homepage

## Optional

- [TED](https://ted.europa.eu/): original source of all award notices