Built for the Clair Z1

mAIday

Describe the analysis you want, in plain English. mAIday assembles it from vetted modules, runs it on your own computer, and saves it as a workflow anyone can re-run.

mAIday means that help is just a sentence away

Type what you need — “run a viability scan on wells A1 to C2”, or “repeat the study in this paper”. mAIday plans the workflow, wires it up, and shows you every intermediate output as it runs.

Z1 connected

You

Run a viability scan on wells A1–C2, then plot dry mass per well.

  • Opened run 21 · 3088×2076
  • Segmented 325 cells · 4 nodes
  • Classified live / dead
  • Dry mass per cell · exported

Segmentation

325 objects

Brightfield image of a well with each of the 325 segmented cells outlined in green

Dry mass per cell

Histogram

Excerpt from the mAIday window, showing real Z1 data.

Assembled from vetted modules

Nothing is improvised. Every step is a tested module written and checked by us — the model only decides how they connect, and you see the result of each one.

  1. Acquire

  2. Load

  3. Segment

  4. Classify

  5. Morphology

  6. Statistics

  7. Export

Modules

Instrument control · Data loading · Metadata · Segmentation · Morphology · Classification · Dry mass and phase metrics · PCA and statistics · Plotting · Export

A worked example

Label-free morphological profiling of a dose–response. Image a 96-well plate, segment every cell, extract forty phase and shape features from each well, reduce and cluster them by phenotype, then fit a dose–response curve per phenotype and export the figures.

  • Written by hand

    2–3 days

    Several hundred lines of Python across five libraries, plus a script to drive the microscope — and only if someone in the group already codes.

The worked example took around 20 minutes to build and refine with mAIday; for an experienced programmer it would be 2–3 days of work, or ~€1,000 of postdoc time.

What it changes

Advanced image analysis, without the tooling. One place to run the instrument, analyse what comes off it, and hand the whole workflow to a colleague.

  1. No learning curve

    No macros, no scripting, no waiting on a bioinformatician. If you can describe the experiment, you can run the analysis.

  2. Measurement to figures

    mAIday drives the Z1 as well as the analysis, so nothing is exported, converted or carried between machines.

  3. Your data never leaves

    Every computation runs on your own machine. Only your instructions and the workflow description reach the language model — never your images.

  4. Reproducible by design

    A workflow is saved as a spec, not a screenshot: send it to a colleague, reload it next year, get the same analysis.

  5. Any assay the module library covers

    The same modules recombine for a new question, so a method nobody in the lab has run is a sentence away rather than a project.

  6. A library that grows

    Modules are written, tested and versioned by us. New ones ship with each release, and you can ask for the one your assay needs.

Experiments that steer themselves

Because mAIday both runs the instrument and analyses the data that comes back, the two can be put in a loop: give it an objective and the experiment adapts while it is still running.

  1. Measure

    on the Z1

  2. Analyse

    on your machine

  3. Compare

    against the goal

  4. Decide

    what to image next

Repeat until the objective is met.

What it costs

  • Lab

    academic · up to 6 users

    €59 / month

    3 new workflows a month

  • Core facility

    academic · up to 15 users

    €149 / month

    10 new workflows a month

  • Industry

    commercial · tailored

    from €399 / month

    Custom allowance and support

  • Re-running a saved workflow is free.
  • Compute runs on your own hardware.
  • All workflows are shared by the group.
  • Extra workflows €20 each.
  • Community workflow database included.
  • Two months free with annual billing.