Know what your AI is fed, what it costs, and what it got right.
Analect is the platform from ETT that shows what your AI is fed, what it costs and whether what it says is supported by its source. Its workflows cover the path from document to checked output: they read and label documents, test whether sending a model less still gets the same answers, price each workload across around 200 models, and certify AI outputs claim by claim with evidence an auditor can check.
Have an account? Sign in to Analect
The company completed two acquisitions in 2023.
ContradictedSource: “The company has never made an acquisition.”
Revenue for the year ending March 2024 was 18.4 million pounds.
SupportedSource: “Revenue for the year ending March 2024 was 18.4 million pounds.”
It employs 240 people across three offices.
SupportedSource: “The company employs 240 people across three offices.”
- Workload
- Contract summaries, 4,000 a month
- Why it won
- Cheapest model that clears the job’s context and quality floor
- Seats or tokens
- Pay-per-use cheaper for the light users
- Priced against
- The catalogue dated 13 Aug 2026
From the document to the checked output
Analect sits beside your AI, not in the path of it. Each workflow measures one decision and leaves a record somebody else can check.
What your AI is fed
What is in the documents your AI reads, who may see them, and does sending a model less still get the same answers?
What it costs
Which model fits each job, what will running it cost, and what will the platform under it cost?
Whether what it says is supported
Does the source support what the output says, what did the output leave out, and can anyone check that we said so?
Start with three weeks on your own bills or your own outputs
Both are delivered by ETT on Analect at a fixed fee. You keep the report and the working file, and nothing commits you to the next step.
AI Spend Forecast
Your AI invoices and workloads turned into model choices, a verdict on seats against tokens, and a budget your finance team can sign off.
- Your current spend explained, by workload and by model
- Every workload priced across around 200 models from 27 providers
- Seats against tokens, with the break-even point
Accuracy Evidence File
A set of your AI outputs checked against the documents they came from, claim by claim, with what was left out reported on its own.
- Every claim supported, contradicted, unsupported or disputed
- Omissions reported separately, never averaged into accuracy
- A certificate for every output, tied to the exact source
For regulated work. The Accuracy Assurance Pack takes the Evidence File over a quarter for one AI system, adds your criteria and your reviewers' decisions, and maps the evidence to ISO/IEC 42001 and the EU AI Act. See the Accuracy Assurance Pack
A number you cannot check is an opinion
Everything Analect produces is written for somebody who has to defend it: a finance committee, an auditor, a client holding a certificate. So every figure says where it came from, and every verdict shows its working.
Every price has a date
A cost without the date of the price list it came from cannot be checked. Every figure in a forecast carries one, and says whether it was measured from your bills or estimated from an assumption we state.
Every verdict has a quote
A claim marked supported or contradicted sits beside the passage of the source it rests on, so a reviewer can follow it back in seconds rather than take our word for it.
Every rate has a denominator
A percentage without its count is a mood, not a measurement. We state what was counted every time: claims checked, claims settled, facts found.
What was left out is counted
An output can be right on every claim it makes and still miss the fact that mattered. Omissions are reported on their own, never averaged into accuracy.
Try the method before you talk to anyone
Are your AI seats paying for themselves?
Compare your ChatGPT, Claude or Gemini seats with the same vendor's pay-per-use prices, group by group, with the break-even point.
Work out your break-even SampleWhat an Evidence File shows
Four AI summaries of one report, checked claim by claim: a contradiction, a true claim the source never makes, and an omission.
Open the sample WhitepaperThe Cost of Asking
Why enterprise AI bills keep rising as token prices fall, and what changing the unit of knowledge does to the arithmetic.
Get the whitepaperHolding a Certificate of Accuracy somebody gave you? Check that it is genuine.
Start with a service. Go further only if it earns it
- Step 1
A fixed-price service
Three weeks on your own bills or outputs, or a quarter for the Assurance Pack. You keep the report and the working file whatever you decide next.
- Step 2
The platform
Where the service shows something worth going further on, your team gets Analect itself: the workflows your work needs, set up by ETT on your own material, with the evidence trail kept as you go.
What we keep, and what we never do
Your bills stay in your browser
The AI Spend Forecast reads your billing files in your browser and never uploads them. We keep the forecast built from them, and you can leave the cost figures out.
Your documents are not kept
The Accuracy Evidence File never keeps your source documents or the outputs themselves. We keep the certificates: the claims, and the quotes each verdict rests on.
Named models, named terms
Checks run on models from Anthropic and OpenAI, through Vercel's AI Gateway, under their commercial API terms. Our sub-processor list is available on request.
On AI cost and AI accuracy
Your AI bill is one line on an invoice. Your finance team needs a budget it can defend.
Why AI spend is so hard to explain, what it would take to run it like any other operating cost, and how we answer it in three weeks.
AI Accuracy · 3 min readIf you sign AI-drafted work, you answer for it. Can you show your working?
What the court cases say about AI output, what proper evidence of accuracy looks like, and how we produce it claim by claim.
AI Assurance · 4 min readISO 42001 and the EU AI Act will ask how you know your AI is accurate. One certificate cannot answer that.
What the two frameworks actually ask for, why the deferred deadline is not a reprieve, and how we turn a quarter of checked outputs into evidence an auditor can use.
Asked before every first call
What is Analect?
Analect is the platform from ETT that shows what your AI is fed, what it costs and whether what it says is supported by its source. Its workflows read and label documents, test whether sending a model less still gets the same answers, price each workload across around 200 models, and certify AI outputs claim by claim. Most customers start with one of two fixed-price services, the AI Spend Forecast or the Accuracy Evidence File. For regulated work, the Accuracy Assurance Pack takes the Evidence File over a quarter for one AI system.
Where does Analect sit in our AI pipeline?
Beside it, not inside it. Analect does not sit in the path of your AI requests, and you do not change your models or tools to use it. Its workflows measure the decisions around a pipeline: what the work should cost and on which model, what is in the documents and who may see them, whether sending a model less still gets the same answers, and whether an output is supported by its source.
Do we need to buy software to start?
No. Each service is delivered by ETT on Analect at a fixed fee: the AI Spend Forecast and the Accuracy Evidence File over three weeks, the Accuracy Assurance Pack over a quarter. Your team gets its own logins to the working file for the length of the engagement.
How do I get access to Analect?
Accounts are set up by ETT for customers, and there is no self sign-up. Each account is given the workflows its work needs and sees only those. If you have an account, sign in at analect.ettgroup.ai/login. If you would like one, book a demo and we will start from what you want to find out.
Does Analect have an API?
Not yet. Today Analect is used in a browser, by account, and what it produces leaves as files: reports as PDF, and estimates, certificates and packs as files your own systems can read. Certificates and packs carry a signature, and anyone can check one on the public verification page without an account.
Where does our data go?
The AI Spend Forecast reads your billing files in your browser and never uploads them. The Accuracy Evidence File does not keep your documents or outputs; it keeps the certificates. Checks run on models from Anthropic and OpenAI under their commercial API terms, and our sub-processor list is available on request.
Does Analect reduce what we spend on AI?
The AI Spend Forecast shows where model choice, caching and seat decisions would change your bill, and by how much, and it says what to test before you move anything. Analect Proof tests on your own documents whether sending a model less still gets the same answers, and reports the input each route used beside the result, so a saving is measured on your material before anybody relies on it.
Can somebody outside our organisation check a certificate?
Yes. Every certificate carries a reference and the digests of the exact source and output it was issued against. Anyone holding one can check at analect.ettgroup.ai/verify that we issued it, that it has not been withdrawn, and that their file is the one it was issued against.
See it on worked material, then on yours
Thirty minutes with the team that runs Analect: the AI Spend Forecast and the Accuracy Evidence File on worked examples, and what either would look like on your bills or your outputs.