← All posts

Recap · 18 November 2026

How the first intake went

Six months, three gatherings and about ten graduates. A short account of how Supratsa's first intake worked and what it taught us

The first intake was small. It began after the Berlin gathering in May 2026, when those who wanted to carry on joined the programme. Then came six months of practice with mentors and two more gatherings.

Berlin, May

Three days under the title “Human in the Loop: Between Versions”. We talked about AI trends and regulation, looked at AI agents in practice, and gave a whole session to AI rollouts that went wrong. The last day was a lab with themed stations and group work.

Lisbon, September

A gathering on digital resilience. Working groups looked at automation at three levels: your own work, the team and the whole organisation, and separately at the new risks it brings. We practised threat modelling on real cases.

Barcelona, October

In the mornings we went to Mozilla Festival sessions: attacks on AI bots, open-source AI on your own servers, small models instead of giant ones. In the afternoons we did our own hands-on work on rolling out AI safely.

What we're taking into the second intake

  • More practice and more time with mentors: that's what people come for
  • Your own task from day one
  • Talking about risks and data as part of every task
  • Six more months of support after the programme

We'll publish the first intake's team stories once the teams have signed them off. You can read more about the gatherings on the Gatherings page.

2027 intake

Apply by 25 December

We'll reply by 10 January

Belarusian NGO or initiative? Please use the Belarusian form

Want to be a mentor? Write to us

Apply by 25 December