How fresh is the schedule?
The file is rebuilt automatically from EMF's official programme, and its timestamp is printed at the top. If your LLM can browse, it's told to pull the live feed and the “now & next” endpoint too — so it stays current as rooms move and self-organised sessions appear during the event.
What does it actually do?
The linked .md is both instructions and data: it tells the LLM to work out who you are (from its memory of you, or a quick optional quiz), then build a tiered, personalised plan from the real EMF programme. It only ever recommends sessions that genuinely exist — no invented talks.
What if it doesn't know anything about me?
Then it runs a short quiz — a few one-line multiple-choice questions about what pulls you to EMF, how hands-on you are, and your rhythm — and tunes from there. Or just say “pick for me” and it'll give you a strong generalist plan.
Is my data going anywhere?
No. Everything happens inside your own chat with your own LLM. This site is static, there's no account, no tracking, and nothing is sent to us — there is no “us” collecting anything.
Is this official?
No — it's a community side-project by an attendee, built on EMF's public schedule API. Huge thanks to the EMF crew for publishing open data. The official schedule is always the source of truth.