I started TenderStack as a bid-writer. I was sick of checking a dozen procurement portals every morning — TED, the national sites, the UK ones — terrified I'd miss the one contract we could actually win. And when I tried ChatGPT, it would happily invent a deadline or miss half the notices.
So I built the tool I wished I'd had: one place that reads every official source itself, matches tenders to what you actually do, and tells you the moment something fits — so you never miss the bid that mattered.
Most companies see a fraction of the contracts they're eligible for — the rest hide in portals they don't check, in languages they don't read, below the thresholds the big aggregators ignore. TenderStack closes that gap.
Searching is the old way: you have to remember to look. TenderStack flips it. Describe what you win once — your services, regions, thresholds, even your go/no-go rules — and it scans every EU, UK and US source on a schedule. The moment a tender matches, it lands in your team's inbox, Teams, Slack or WhatsApp, already scored and screened. You react to opportunities instead of hunting for them.
Run several watches at once — one per service line or country — each with its own criteria and its own channel. It's the difference between a tool you open and a teammate that tells you when it matters.
Beyond search, TenderStack ships expert skills that activate inside Claude and ChatGPT — so the same assistant that found the tender can help you research, qualify and plan the bid. These run in the chat, not the web app.
Public procurement is governed by statute, regulation and a deep body of case law — and the rules differ in every country. The Legal Assistant module reads the live tender against the actual law of its jurisdiction and flags issues before they cost you the bid: disproportionate selection criteria, unlawful eligibility requirements, missing motivation, deadlines that breach the directive. Every flag carries a citation to the primary source — never an invented rule.
A continuously-refreshed legal layer over official primary sources — statutes, EU directives, contract law, advisory opinions and the decisions of national procurement-review bodies and the Court of Justice of the EU.
Coverage spans the EU (incl. NL, DE, FR, ES, IT, BE, AT, IE, PL, CZ, DK, SE, FI, GR, RO, SI, EE, HR, CY, MT), the UK and the US, and grows over time. The Legal Assistant is informational support — it surfaces and cites the law; it is not a substitute for advice from a qualified lawyer.
No scraping, no stale third-party feeds, no black box. TenderStack reads directly from official public registers and open data — refreshed every 8 hours, with history back to 2023 — so the intelligence you bid on is current and verifiable.
Real output from TenderStack running inside Claude. No demos, no mockups.
One API key, two platforms. Pick whichever AI you already use.
sk_tdb_… key.https://mcp.tenderstack.io/sse?key=sk_tdb_…sk_tdb_… key when prompted. The same 19 skills and 14 data sources are available.Start free, no card required. Upgrade when TenderStack is already earning its keep. Cancel anytime.
Statutes, EU directives and procurement case law for 27 jurisdictions — a cited 🟢/🟡/🔴 legal read on any live tender, inside Claude, ChatGPT or the web app. Add to any plan.
TenderStack isn't just for humans. Autonomous agents and LLMs browsing the web can query live tenders, awards and grants — and act on them — over MCP or a plain REST API, no scraping required.
TenderStack is built by LemonB BV, a certified tender-management company with 10+ years of hands-on experience winning bids across Europe and the US. We've run hundreds of procurement cycles — from pre-solicitation strategy through award — and encoded that expertise into the product.
TenderStack is the open front door to our wider platform, TenderB — and the two play very different roles:
Find, qualify and draft against every public tender across the EU, UK & US — from inside Claude, ChatGPT or any AI agent. Discovery, who-won data, scouting and proposal drafting, API-first. Nothing to log into.
The complete bid-management brain — a modular workspace where your team runs the entire tender lifecycle: every source ingested, a collaborative AI writing canvas, a configurable workflow engine, jurisdiction Skills (NL · DE · UK), approvals, audit trail and memory that learns your win/loss patterns.
Visit TenderB ↗Start in TenderStack to discover and act on opportunities today; graduate to the full TenderB workspace when you're ready to run bids end-to-end.
A general-purpose model writes fluent prose — and will confidently invent the regulation, threshold or citation that gets you disqualified. Winning public tenders isn't a writing problem; it's a compliance, evidence and memory problem. That's the part a raw chatbot can't do — and exactly what TenderB is built for:
TenderB's NL · DE · UK Skills are curated and continuously evaluated against real procurement law, with citations to source. A generic model hallucinates the rule it can't cite.
It writes in your organisation's voice, reuses your proven, won answers, and learns your win/loss patterns. ChatGPT starts from a blank page every single time.
EU data residency, a full audit trail and ISO 27001. You cannot drop a confidential RFP into a consumer chatbot and stay compliant — your bid team can't risk it.
Every customer-bound output passes hallucination filters and human-in-the-loop checkpoints. No surprises in front of an evaluator who can disqualify you for one wrong claim.
TenderStack puts live opportunity intelligence in the chatbot you already use. TenderB is where that intelligence becomes a compliant, on-brand, evidence-backed bid you can actually submit.
Connect TenderStack. In Claude, add the MCP server at https://mcp.tenderstack.io/sse; in ChatGPT, use the TenderStack GPT. You can then ask in plain English for live public tenders, contract awards and grants across the EU, UK and US — and get AI summaries, who-won data and re-tender forecasts.
TenderStack is an AI-native public-procurement tool that works inside Claude, ChatGPT and a web app. It covers every public tender across the EU (all 27 member states via TED), the UK, Netherlands, Belgium, Germany, France and the US — above and below threshold — with semantic search, so you match on what your company does rather than memorising CPV codes.
Yes. TenderStack includes below-threshold tenders for Germany, the Netherlands, Belgium, the UK, France and the US, plus all above-threshold EU tenders via TED — over 1.5 million notices, refreshed every 8 hours, with history back to 2023.
TenderStack stores award data on closed tenders, including the winning supplier and award value where published, so you can analyse incumbents and price to win.
Yes. TenderStack has a free plan with 200 actions per month; Pro ($49/month) is unlimited. It works in Claude, ChatGPT and the web app at app.tenderstack.io.
EU-wide via TED (all 27 member states), the UK (Find a Tender and Contracts Finder), the Netherlands (TenderNed), Belgium (e-Procurement/BDA), Germany (oeffentlichevergabe.de), France (BOAMP), and the US (SAM.gov, Grants.gov, USASpending, FPDS and more).