Techmirionτεκμήριον · proof

Private AI for the work you can’t afford to get wrong. It runs on your own machines, gives the same answer to the same question, and shows where each result came from.

Talk to usRead the research
  • PublishedarXiv:2605.13245
  • Used dailyat IESL-FORTH, Heraklion
  • Same answer4 runs out of 4
  • Weeks of workdown to minutes
peak_analysis v1.4·lab-sample.csv
range 530–625 nm · try to drag it
measurement our result rangeIllustration · made-up sample data

From our researchTable 3

4/ 4

runs gave the same answer with our tool. Three leading AI platforms: 0 of 4.

See the full test

01/The problem

Most AI makes you choose between private and powerful.

The most capable AI runs in someone else’s cloud, so using it means sending them your data. The private kind, on your own machines, is usually the weaker one. Most teams end up giving up one or the other.

We build AI that stays on your machines without falling behind. It works best on the repeatable processes your work depends on, once we’ve written down how they’re done.

The usual choice
Privacy030
Performance070
Their cloudYour machines

Your data stays with you. You don’t give up performance for it.

02/How it works

AI that follows your method, every time.

Ask a chatbot the same question twice and it may take two different approaches. That’s fine for a first draft. It’s a problem when the answer is a measurement, a deadline, or a number someone has to sign off.

So we don’t let the AI decide how the work gets done. We sit down with the people who do it and write down their steps, settings and rules. Each step becomes a tool that works the same way every time. The AI only picks the tool and fills in the details. If a request breaks the rules, it’s stopped before it runs.

Example · how a request gets checked

  1. 01AIpicks the tool and its settings
  2. 02Checkstops a request that breaks the rules
  3. 03Tooldoes the work the same way every time
  4. 04Receiptshows where the result came from

What the AI asked for

{  "tool": "peak_analysis",  "range_nm": [520, 610],  "peaks": 1,  "baseline": "straight"}

Waiting for a request…

  • Consistent
    Same question, same answer. Run it again tomorrow and you get the same result, down to the last digit.
  • Sourced
    Every result shows where it came from: which data, which step, which rule.
  • Tested
    We score it against answers you already trust, on cases it has never seen. You get the score with the system.

When a better AI model comes out, we can swap it in. Your method stays as it is.

03/The test

Four tries each. Only our tool gave the same answer every time.

We took one real measurement from the lab and gave it to our tool and to three leading AI platforms, all on paid plans with every feature turned on. Each one got the same file and the same plain-English request, four times.

  • Techmirionour tool
    1234
    –/ 4–/ 4

    The same result in all four runs, 34 seconds each.

  • GPT-5.5
    1234
    –/ 4–/ 4

    Saw one peak in some runs and two in others. Its results moved by up to 54.8% from run to run.

  • Claude Sonnet 4.6
    1234
    –/ 4–/ 4

    Measured the wrong feature in most runs, and drew its boundaries differently each time.

  • Gemini 3.1 Pro
    1234
    –/ 4–/ 4

    Got a unit conversion wrong every time, so none of its results were usable.

  • usable, and the same answer every run
  • usable, but the answer changed between runs
  • not usable

This test measures consistency. Choosing the right approach is the expert’s job, so we build theirs in.

SourceAdamidis, Katrisioti, Tzitzikas, Stratakis. arXiv:2605.13245, Table 3. May 2026.

04/In use today

Already running, in research and in law.

AResearch labs

FORTHought at IESL-FORTH

FORTHought is the AI platform we built for the Stratakis Lab at IESL-FORTH. Researchers use it every day to search their papers, analyse data and process results from their instruments. It runs on the lab’s own servers, and nothing they send leaves the building.

Spectrum analysis~2 weeksminutes

per batch of measurements · closely matches the researcher’s own results

Electron-microscope pattern analysis~2 daysminutes

checked against calibration · used by three researchers

11regular users over six months, when the paper came out

BLaw

A case-file assistant for Greek lawyers

It reads the case file and helps with the work around it.

  1. 01Works out the right procedure and the right court, following the rules.
  2. 02Calculates deadlines and explains each one. If it can’t calculate one, it says so and why.
  3. 03Finds Supreme Court rulings for and against your client, most relevant first.
  4. 04Drafts the court filing and marks missing information in red, instead of making it up.

It runs on our own servers in Heraklion. No document is sent to an outside company.

60,000+Supreme Court (Άρειος Πάγος) rulings
~60codes, laws and EU acts, up to date as of Law 5326/2026

Figures as of 16 Oct 2026

CNext

Your process

Anything that has to run the same way every time, on data that should stay in-house. Lab analysis, case files, compliance checks, technical reviews.

Tell us about it

05/Working with us

Four steps from the first call to a tested system.

  1. 01 /

    Listen

    We sit down with the people who do the work and, over a few short sessions, write down how they do it. On our first project, 7 of their 30 answers changed what we had assumed, so we never skip this step.

    7 / 30answers that corrected our first version

  2. 02 /

    Build

    We turn each step into a tool that always works the same way. The AI decides when to use each tool, but never changes how the work is done.

  3. 03 /

    Install

    We set it up on your own machines, next to your data and the software you already use. Nothing leaves your building.

  4. 04 /

    Test

    We score it against results you already trust, on cases it has never seen, and show you the number with its margin of error. You see it before you commit.

After delivery
A member of our team stays on to look after the system and answer your questions. When better AI models come out, you get them, and your method stays the same.

06/Who owns what

What you own, and what we keep.

Yours

  • the system installed on your machines
  • your data
  • your methods and procedures
  • every result it produces

Ours

The technology we use to build, tune and test your system. It’s what lets us keep improving it after delivery.

07/Who it’s for

Built for teams whose data can’t leave the building, and whose answers have to stand up to checking.

  • 01research institutes and labs
  • 02law firms and bar associations
  • 03engineering and technical teams
  • 04regulated businesses
Not the right fit
Early exploring, where nobody has settled on a method yet, or tasks where a rough answer is good enough. A general chatbot will do fine there, and we’ll tell you so.

08/Our principles

What you can hold us to.

  • 01
    Every number has a source. If we can’t show where a figure came from, we don’t publish it.
  • 02
    Research comes first. Our products grow out of our research. When the two pull in different directions, the research wins.
  • 03
    We show our evidence. Our claims point to published research you can read and check for yourself.
  • 04
    We’re upfront about limits. Our paper lists what it doesn’t prove, and the report you get from us will do the same.
  • 05
    We don’t train on your data. Nothing you send is used to train an AI model.
  • 06
    We take our machines back. When they reach the end of their life, the machines we install come back to us to be refurbished, recycled or rebuilt.

09/Our name

Techmirion comes from τεκμήριον (tekmírion), the ancient Greek word for proof: the evidence that settles a question. We chose it because proof is what we deliver.

Our logo is built on the small filled square that mathematicians put at the end of a proof. You’ll see it on this page wherever we’ve shown ours.

claim
same question, same answer
source
arXiv:2605.13245, Table 3
checked
4 runs, the same result
The square that closes a proof

τεκμήριον — proof. We’d rather show you the evidence than ask for your trust.

10/Who we are

Built by the researchers who use it.

Techmirion is the company arm of the Stratakis Lab at the Institute of Electronic Structure and Laser (IESL) of FORTH, in Heraklion, Crete. FORTH is one of the largest research centres in Greece.

We’re a small team of researchers and engineers. We built the lab’s AI platform, we run our own AI model on our own servers, and we rely on both every day. What we offer you is what we already use ourselves.

We work to the institute’s scientific standards, as an independent company.

Supported by
Dr. Emmanuel Stratakis, Director of IESL-FORTH, and Prof. Yannis Tzitzikas of the Computer Science Department, University of Crete, and FORTH-ICS.

11/Research

Read the research behind it.

PreprintarXiv:2605.13245 · May 2026

“It’s not the Language Model, it’s the Tool: Deterministic Mediation for Scientific Workflows.”

M. Adamidis, D. Katrisioti, Y. Tzitzikas, E. Stratakis

What we’re working on next

  1. 01a formal study with users
  2. 02support for more lab instruments
  3. 03systems at more organisations

Get in touch

Start with one process.

Pick the one that takes the most time, or would do the most damage if it went wrong. We’ll tell you plainly whether we can help, and what results to expect.

Email ushello@techmirion.com