The AI researcher index
We read the papers and the public signals that say when a PhD ends. You see who does the best work on your subject, and you hear months ahead when they are about to be free, while they are still reachable.
The index and the alerts
Search any subject in your own words and see who leads the work on it, whether they are ten years into a career or about to defend. Set an alert and hear before anyone else when one of them is about to be free.
Access the index, $179 a month Pay by card and start in a minute. See pricing.The recruitment agency
We find, approach and shortlist the best AI researchers for you, from the same index. You pay only when one of them joins your company.
Book a call No retainer, one fee on a signed offer. How it works.Why it works
Some roles need the person who already leads the work on your subject. The index ranks people by what they published and where it appeared, so you write to the five who matter rather than the fifty who used the keyword.
Other roles need someone who is about to arrive, and there timing is most of the job. A researcher is reachable during the months around finishing a PhD, and after that they are inside a frontier lab on a package you cannot match. The index dates the end of a doctorate from public records and tells you months ahead.
Type the subject in your own words. We match the title and abstract of every indexed paper, so you reach the people doing that work today, at whatever stage of their career, rather than the ones who wrote the right keyword on a profile.
Each profile leads with what they published, where, and with whom. You judge the work first, which is also what makes the first message worth answering.
An alert lands when a new researcher appears on your subject, and again when their date firms up. That is the moment to write, and it comes months before a job board would tell you anything.
In practice
Four steps, and the fourth is the one that gets you the hire. Every date on a profile carries the source we read it from, so you know what you are acting on.
The subject in your own words, the way you would describe the role to a colleague. No preset categories to choose from.
The people whose published work is closest to it, with what they wrote, where, and with whom.
When they finish and come on the market, months ahead, and where that date comes from.
An alert tells you when a date firms up. You reach them about their work, before a lab does and before a package you cannot match.
The people you want were not looking, and nobody had written to them yet. Arriving before that changes is the job.
Founding Hires, AI research and engineering search
The recruitment agency
Same index, our hands. We find the people, approach them about their work, and hand you a short list with our reasoning attached. You pay only when one of them joins your company.
Built from arXiv, OpenAlex, ORCID and DBLP, rebuilt every night, and measured rather than estimated. The method page says how each field is filled and where it stops being reliable.
| What the index holds | Count |
|---|---|
| Papers indexed | 554 206 |
| Researchers on those papers | 1 027 515 |
| Profiles enriched from OpenAlex, ORCID and DBLP | 1 018 926 |
| Profiles with a classified research area | 990 696 |
Updated 2026-10-14
The engineering half
A paper tells you who solved the problem. It seldom names whoever shipped the training loop, the inference server or the evaluation harness. Those people leave their trace in commits instead. We built a second tool for them: StarHunt, GitHub sourcing for AI engineers.
It reads GitHub in real time and filters on stack, recent activity and the kind of AI work people commit, so you reach engineers no keyword search on a job board will ever surface. Same principle as the index, applied to code instead of papers, and open to you directly.