Check they have actually done this, before you spend an afternoon on it
Vetting a speaker properly takes half a day. Doing it to someone who has never taken a stage takes the same half day. We index the agendas of 39 European conferences so the cheap check comes first.
Get early access
Free during the beta. We let agencies in ten at a time so we can actually talk to each one.
What is in the directory today
- speakers who took a stage
- 1200+
- European events, 15 countries
- 39
- linked to a public agenda
- 100%
Three entries, picked at build time
Not testimonials and not logos — we have neither, and a page about evidence is the wrong place to invent them. These are records from the dataset. Open one and check it against the conference's own page.
“Demo 1 - Forward Deployment at OpenAI”
Ilyass El Mansouri
AI Summit Barcelona 2026
verifiedsource · aisummitbarcelona.com“Beyond LLM-as-Judge: Using GLIDE for Reliable, Scalable Evaluation of GenAI Systems”
Grégoire Martinon
PyData Amsterdam 2026
verifiedsource · pretalx.com“Building the Nervous System of AI”
Michael Kagan
WeAreDevelopers World Congress 2026
verifiedsource · wearedevelopers.com
The order most vetting advice gets wrong
Nothing in it is bad advice. It is just the expensive half, done first.
What you are told to do
Watch twenty minutes of footage. Check for a speaking credential. Phone two organizers who booked them before. Every bureau’s vetting guide says some version of this, and none of it is wrong.
Why it is the wrong first step
It is an afternoon per name, and you are spending it before you know whether the person has stood on a stage at all. The expensive check comes first and the cheap one never happens.
What to check first
Whether they spoke, where, when, and on what — from the conference’s own agenda rather than from a profile they wrote. It takes seconds. Then spend the afternoon on the two who survive it.
That advice is everywhere for a reason: it is what you are left with when the only record of a speaker is the one they wrote about themselves. A directory people pay to join cannot tell you anything else. See how to find speakers for the step before this one, or how the dataset is built.
What this does not cover
Stated here rather than discovered later. A directory arguing for evidence cannot be vague about its own edges.
AI and adjacent topics, not every subject
The dataset is built from AI and tech conference agendas. If you are booking outside that, this will not help you yet.
Europe. The US is not covered
39 events across 15 countries, all European. US coverage: not yet.
We rarely know how to reach them
The agendas we source from almost never publish contact details, so only a small fraction of entries carry a link. We show you who spoke and where; getting in touch is still your step.
Reasonable objections
The four we hear most, answered without the marketing.
- Why not just ask ChatGPT?
- You can, and it is free. Ask it where someone spoke last quarter, then click one of our source links and compare. One of us is reading an agenda and the other is predicting a plausible sentence.
- LinkedIn already works for me.
- It does, at about a day per shortlist. This is built to get you to the right ten profiles instead of the wrong hundred — you still land on LinkedIn at the end.
- A bureau does this for me.
- For a €25k keynote, keep your bureau. For the six practitioner slots underneath it, bureaus do not play in that tier and this is all we do.
- Is this GDPR-clean?
- Public conference agendas only. We never touch LinkedIn’s data, every entry names the source it came from, and anyone can ask to be removed — the removal survives the next rebuild.
Check one yourself first
The directory is free to read and there is nothing to sign up for. Pick any speaker, follow the source link to the conference’s own agenda, and see whether it says what we say it says.