Ask about a buyer, a competitor or a market in plain language, in the AI tool you already use. Tenderlake Analyst answers from more than a decade of public procurement history, and shows you the evidence behind every claim.
Someone pulls the notices, tallies the awards, works out who is actually winning, and writes it up. By the time it reaches the people who asked, the meeting has happened and the business has moved to the next priority.
So research stops being something you do when a question arises and becomes something you diarise. Quarterly reviews. Annual planning. The three biggest pursuits. Every other decision gets made on instinct and whatever someone half-remembers.
Tenderlake Analyst removes the lag, and with it the reason to ration the questions.
Describe what you want to understand in plain language: the buyer, the competitor, the market, the route in. No exports, no spreadsheets, no query syntax. Tenderlake Analyst works inside the AI tool your team already uses, on data that goes well beyond the standard Tenderlake interface.
Every claim comes with the notices, awards, dates, values and named buyers behind it. Tenderlake Analyst separates what the data proves from what it suggests, and you can ask it directly which findings are evidenced and which are inference, before you build a pursuit decision on them.
The first answer is the start. Push on it, narrow it, challenge it. The conversation remembers what it found. Then turn the same analysis into a management memo, a set of meeting questions or a six-slide account briefing without leaving it.
| Manual research | Tenderlake Analyst | |
|---|---|---|
| How you ask | Exports, filters, spreadsheets | A question in plain language |
| What you get back | Rows to interpret | Analysis with named buyers and suppliers |
| Time to an answer | Days | Minutes |
| Evidence | You assemble it | Notices, dates, values and buyers cited |
| Follow-up questions | Start again | Same conversation, builds on what it found |
| The output | You write it up | Memo, meeting questions or deck on request |
| Languages | Manual, one at a time | Translated across EU languages |
More than a decade of awards, notices and forward plans, translated across EU languages and name-matched so one buyer reads as one buyer however it was written down. Tenderlake Analyst searches it by meaning, not by keyword.
What they have awarded, to whom and at what values, and what their pre-procurement and planning notices signal about the next twelve months.
See what a named competitor has won, from which buyers and in which categories. The same view shows where their position is thinnest, and which buyers no one has locked down.
Where demand for the kind of work you sell is strongest, which buyer types are driving it, and whether the spend is growing, flat or consolidating into fewer suppliers.
Whether they buy through frameworks, dynamic purchasing, direct award or open competition. Which of those routes carry the real spend in your space, and whether a given framework is worth the effort of getting on.
Ask a general model about a public buyer and it will answer, fluently, from whatever it half-remembers. It has no award records, no notice history, and no way to tell you where it is guessing. Tenderlake Analyst gives the reasoning something real to stand on.
Claude · ChatGPT · Gemini · Copilot
and most major AI assistants
Tenderlake Analyst is a remote MCP connector. Nothing is installed, and no API keys are ever handled by your users.
Tenderlake Analyst is used by the people who carry a public sector growth number. Different questions, the same working pattern.
Walk into every buyer meeting informed. See which suppliers hold the work inside an account you already own, and find the gaps before a competitor does.
Decide where a small team places its time, with evidence for which buyers are active, growing and winnable.
Understand how a buyer actually buys, who holds the work now, and when it comes round again.
Hold a defensible view on where a market is heading, without a research team to build it.
Take market entry and investment decisions to a board with the evidence behind them.
See which buyers are spending on the propositions you are taking to market, and which suppliers already sit alongside you in those accounts. Size a partner or a reseller by the public work they actually hold, not the work they claim, and know where a joint offer opens a door that neither side opens on its own.
The Tenderlake Analyst brochure covers what it answers, the data underneath it and how it is set up – in a form you can forward.
Your Tenderlake Workgroup administrator enables Tenderlake Analyst for your workgroup, and buys credits. A paid subscription to Tenderlake Monitor is required.
Access is off by default. Your administrator grants it per colleague, and can set a monthly credit cap at the same time.
In Claude: Settings, Connectors, Add custom connector. They sign in with their own Tenderlake login and approve. No keys or codes to copy.
That understanding used to take a week to assemble, so it only got built for the biggest decisions.
Tenderlake Analyst is not a faster way to do research. It is a reason to ask more questions.
Book a demo of Tenderlake Analyst Download the brochureTenderlake Analyst is a connector that lets the AI assistant your team already uses read Tenderlake’s procurement record and analyse it for you. You ask a question in plain language – about a buyer, a competitor or a market – inside Claude, ChatGPT or whichever assistant you use, and the answer comes back built on award records, notices and forward plans rather than on whatever the model happens to remember. It adds a governed data source to an assistant you already control, and it reaches data that goes well beyond the standard Tenderlake interface.
Ask a general model about a public buyer and it will answer, fluently, from whatever it half-remembers. It has no award records, no notice history, and no way to tell you where it is guessing. Tenderlake Analyst gives that reasoning something real to stand on: every claim comes back with the notices, awards, dates, values and named buyers behind it, so you can check it before you act on it.
Claude, ChatGPT, Gemini, Copilot and most major AI assistants. Analyst is a remote MCP connector, so any assistant that supports custom connectors can use it. The requirements on your side are that your assistant plan permits custom connectors, and that your own AI governance policy allows this one.
Yes. Tenderlake Analyst runs alongside a paid Tenderlake Monitor subscription rather than replacing it, and it is not available on the free trial. It is metered separately in credits, which your workgroup administrator buys. See Tenderlake Monitor pricing for the subscription itself.
Four kinds. Know the buyer: what they have awarded, to whom and at what values, and what their planning notices signal about the next twelve months. Know the competition: what a named competitor has won, from which buyers and in which categories, and where their position is thinnest. Read the market: where demand for the kind of work you sell is strongest, and whether spend is growing, flat or consolidating into fewer suppliers. Find the route in: whether a buyer buys through frameworks, dynamic purchasing, direct award or open competition, and whether a given framework is worth the effort of getting on.
The first answer is the start, not the end. The second and third questions, asked in the same conversation, are usually where the value is.
More than a decade of awards, notices and forward plans, covering over two million organisations. Buyers and suppliers are name-matched, so one buyer reads as one buyer however it was written down, and Analyst searches the record by meaning rather than by keyword. What you can reach is bounded by your own plan’s regions.
Yes. The record is translated across EU languages, and a competitor’s wins are matched by company name whatever language the notice was published in. Sizing up a supplier’s position in three countries is one question, not three questions in three languages.
Yes. Answers cite the notices, awards, dates, values and named buyers they rest on, which is what makes them usable in a pursuit review or a board paper. You can follow up in the same conversation to pull any single claim back to the record it came from.
Yes, and it is worth asking. Analyst separates what the data proves from what it suggests, and you can ask it directly which findings are evidenced and which are inference, and to flag anything to verify before it goes outside your organisation.
One practical caveat: company and buyer names vary between notices. If a firm or a buyer looks missing, ask it to try other spellings and trading names.
Your Tenderlake workgroup administrator enables Analyst for the workgroup and buys credits, then grants access to individual colleagues. Each colleague adds a custom connector in their own AI assistant, signs in with their usual Tenderlake login and approves it – about a minute per person, with no keys or codes to copy.
One thing to know afterwards: they need to start a new chat, because connector tools are not attached to conversations that already exist.
No. Analyst is a remote connector reached over HTTPS. Nothing is installed on user devices or on your network, no new AI model is introduced, and your users never handle an API key – they sign in on Tenderlake’s own login page.
No. Access is off by default and granted per person by your own workgroup administrator, who can revoke it at any time. A colleague who has not been granted access is told to ask their administrator, and no credentials are issued.
It is metered in credits rather than sold as another monthly seat, so you pay for what you ask. You are charged for two things: each search your assistant runs, and each row of the procurement record it brings back. The assistant’s thinking is not metered.
Questions about your own folders and saved searches are not metered at all – they are included in your Tenderlake subscription – and setting up the buyers and competitors you track is free. You are charged when you reach into the wider market. Ask us for current credit prices.
Yes. Your workgroup shares one credit balance, and your administrator can set a monthly credit cap per person, with an email alert when somebody is close to their limit. A cap of zero stops that person’s Analyst calls without removing their access, and caps reset at the start of each month.
Submit your question here and we will get right back to you.