Table of Contents
← Back to the Final Project overview
You have a dataset and a question you have never had time to answer, and a student on the Data-Driven Life Sciences course (KTH / SciLifeLab) is going to take it on. They direct an AI agent to do the analysis and check every result themselves; you supply the problem and the judgement of whether the answer is useful. This page is everything that matters from your side.
What we ask of you
About two hours, total. Two one-hour online meetings, roughly three weeks apart. Everything between and after them happens without you.
- It is free, and there is no obligation — you can stop at any point.
- The results are yours. We make no claim on your data, your results, or authorship.
- At the end you get a written report and a small analysis app or the code, which is yours to keep.
How it works
| Step | Your time | What happens |
|---|---|---|
| 1. Matched | none | We pair you with a student and confirm the project. You get a private link to your project page. |
| 2. Meeting 1 | 1 hour, online | The student interviews you about the question — and this is where you hand over the data (see below). |
| 3. They build | none | Two or three days. The student directs an AI agent to do the analysis and checks every result. |
| 4. Meeting 2 | 1 hour, online | They show you what they built; you tell them what is wrong. Being corrected here is the point, not a failure. |
| 5. Optional rounds | by agreement | Only if you both want to — never assumed. |
| 6. You get the results | none | A short written report plus a runnable app or code. The student hands their work to the course for assessment, and we share the results back to you. |
How your data is handled
Your data is yours, and we treat it that way. You choose how it is shared, at Meeting 1:
- A copy or an extract. A de-identified or subsampled extract is usually plenty — you do not have to hand over everything.
- A protected environment (the default for anything sensitive). We run a small, locked workspace on a course server (KTH SciLifeLab). You upload your data into it; the student’s agent can run analysis inside and see results, but cannot download or copy the raw data out. Every command and output is checked. When the project ends, the workspace is torn down.
- On your own machine. If you would rather not move the data at all, the student can design the analysis and you run it yourself. That is a perfectly good choice.
We never move or upload data you have not explicitly agreed to move.
Questions about protecting a specific dataset, or setting one up: email joanna.hard@scilifelab.se.
How you follow along
You have your own private project page — a link we email you, with no account or password. It is the single place the whole project lives, so nothing gets lost in email. There you can:
- upload your data and a sample file — the tidiest way to hand it over;
- watch the stage move — matched → meeting 1 → building → meeting 2 → delivered;
- read the student’s progress updates and reports, including the final one;
- raise a question or an issue in a thread the student and the teaching team both see;
- get the occasional email on the big moments — which you can turn off.
Lost your link? Just email us and we will re-send it.
Your dates
| When | What |
|---|---|
| Late Sep – early Oct | We match you with a student and confirm the project. |
| Early–mid Oct | Meeting 1 — interview and data hand-over. |
| Mid Oct | Meeting 2 — you see the prototype and say what is off. |
| From ~23 Oct | You receive the final report and the runnable app/code. |
(The student works to a 23 October hand-over; your results follow shortly after.)
Questions
Email the teaching team at ddls-course@scilifelab.se any time. For the fuller background on what the course is and why we work this way, see the original call for projects.