Getting started with Torhy
What Torhy does, how to connect it to your AI assistant and run your first tender search in five minutes.
Torhy gives your AI assistant (Claude, ChatGPT, Cursor, Claude Code) access to open public procurement data: Ukrainian tenders from Prozorro and EU notices from TED. You describe the task in plain language and the assistant calls the right Torhy tools and explains the result.
What you can do
- Search open tenders by keywords, CPV (ДК 021) codes, region, value, procedure type, dates and buyer.
- Read tender documents — text from PDF, DOCX, DOC, XLSX, RTF, ZIP and e-signed (.p7s) files, split into pages of about 12,000 characters.
- Analyse the market — completed procurements with winners, discounts and number of bidders, unit prices (median and quartiles), buyer and supplier profiles.
- Get alerts about new tenders matching saved searches and about changes in tenders you watch.
- Search EU tenders on TED by country, CPV code and keywords.
Three steps
- Sign up at https://torhy.com.ua (passwordless email link).
- Add the connector with the URL
https://torhy.com.ua/mcp— guides for Claude, ChatGPT and Cursor. - Ask your first question, for example:
Find open tenders for roof repairs in Kyiv Oblast with an expected value below UAH 2 million.
Company profile
Save your company profile — CPV codes, regions, keywords, value range and a capabilities summary. Then simply ask “Find tenders for my company” and the assistant applies the profile filters.
Who does what
Torhy supplies fresh structured data, document text, analytics and alerts. Judgements — whether a tender fits, which conditions are risky, what to ask the buyer — are made by your assistant. Always check important conditions in the original documents on Prozorro or TED; every result links to the source.