Want an AI agent to do your analysis? Bring a laptop.
December 3, 2026 @ 09:00 – 17:00 CET
One that keeps working after you have gone home, that your collaborators can ask too – and that never moves your data off your disk. A hands-on day with Mycolab.
Four questions. If you say yes to any one of them, this day is for you.
– Do you want an AI agent to do the analysis for you? On the files already on your disk – not on a copy you uploaded somewhere.
– Do you want your whole team working on the same data? In one place, instead of mailing versions of a notebook around and losing track of which one made the figure.
– Are you tired of sending a collaborator 40 GB and then spending a month in meetings explaining which cells count? You export a figure, they mark it up, you re-run it, and again. Every lap moves gigabytes and costs a week.
– Is there sensitive data in your lab sitting untouched – waiting months for the approvals that would let an external expert near it – because the only way to get their help today is to hand them a copy?
Those are four versions of one problem: the data has to move before anyone can help you.
This day is about the other way round.
What Mycolab does
It puts an AI agent next to your data – on your own laptop, or on the server your files already sit on – working under rules you write once. You ask it to do the work. In this lab that means segmenting organoids, tracking them through a time series, flagging the ones that divide; it is the same shape whatever your files are, and nothing in Mycolab is specific to images. A collaborator can ask too, from wherever they are. You see every request, you approve what leaves, and nothing else does. The data never goes anywhere, and the agent keeps working after you have gone home.
Spend the day with us and you leave with that running on your own machine. Not a demo you watched, not an account you opened, not a pilot you were promised.
Morning – presentations
What this is, what it is for, and what each piece does. Speakers: Wei Ouyang, Hanzhao Zhang, Joanna Hård, Nils Mechtel
Afternoon – hands-on, all of it.
Mycolab set up on each person’s own machine, then working with it on real files. The day closes with a demo.
There are 25 places, and the day is free.
Participants must bring their own laptop.
By the end of the day you will have:
– your own agent running in the folder your data is already in, working on real files – bring your own, of any kind, or use one of our example microscopy datasets;
– a dataset that colleagues can find and ask questions of, announced rather than uploaded – it stays on your storage, in your format, under your licence. We practise this on the FAIR Microscopy Data Hub; the pattern is the point, not the file type;
– seen bioimage analysis run where the data lives, with BioEngine;
– seen Safe Colab, our work on letting a collaborator analyse data they are never allowed to see – the analysis goes to the data, and only the result comes back.
Who it is for
Anyone who makes more data than they can analyse or share, in any field – PIs, postdocs, PhD students, analysts, data stewards and core facility staff. Microscopy is where this lab works, so our examples come from there, but nothing in Mycolab assumes your data is images. No previous experience with AI agents is needed.
What to bring. A laptop is required – the afternoon is hands-on and we install Mycolab on your own machine. If your data sits on a lab server you can reach, bring that connection too. You do not need to bring data of your own – there are example datasets – but if you do, you will leave having done something real with it.
The day is free.
A minute and a half, if you would rather watch than read: https://tl-mycolab-film-38275b5b.hypha.run/
CONTACT FOR QUESTIONS
Hanzhao Zhang – hanzhao.zhang@scilifelab.se

