Guide
The best tools to find AI researchers, compared
Five sourcing tools for AI researchers and AI engineers: what each one reads, whether it knows one person from two, whether it links to the work, and whether it can tell you when someone becomes available.
The short answer
The best tool to find AI researchers is one that reads the research itself, not what people write about themselves. The Founding Hires index holds 1 027 515 AI researchers on 554 206 papers, one profile per person, with every paper linked and, for 19 781 of them, the end of the doctorate dated from a public source. A generalist network such as LinkedIn sees the people who keep a profile up to date, and most researchers do not.
The same index covers applied AI engineers and AI engineers, who sign papers as middle authors or publish models. For software engineers, StarHunt reads GitHub and ranks people on what they commit. Google Scholar is free and good for reading one researcher's record, not for sourcing many. LinkedIn Recruiter covers every profession and every market. SeekOut aggregates public profiles for technical recruiting at scale.
| Founding Hires | StarHunt | LinkedIn Recruiter | Google Scholar | SeekOut | |
|---|---|---|---|---|---|
| Built for | AI researchers, applied AI engineers and AI engineers | Software engineers, from their public code | Every profession, every market | Reading one researcher's publications and citations | Technical recruiting at scale |
| What it reads | The published AI literature: arXiv, OpenAlex, DBLP, ORCID, proceedings, theses | Public GitHub activity, read in real time | Profiles written by their members | Papers found on the web, and profiles created by researchers | Public professional profiles aggregated from several sources |
| One profile per person | Yes: name variants merged into one researcher (OpenAlex, ORCID) | One profile per GitHub account | One profile per member, as they wrote it | Only for researchers who created a profile; other author names stay unlinked | Merged records; ask how duplicates are resolved |
| Linked to the work | Every paper, venue and author position, linked to the source | Repositories, commits and the stack they use | What the member chose to list | Papers and citation counts | Depends on the source of each record |
| University and lab | Affiliation read on the latest paper, across 132 countries | Location and organisation declared on GitHub | Declared by the member | Declared on the profile, when there is one | From the aggregated profiles |
| When they become available | End of doctorate, with its evidence: thesis, public statement or estimate | Recent activity and open-to-work signals on GitHub | Open-to-work flag, when the member sets it | Not tracked | Not tracked for doctorates |
| Price | $179 a month for the index, with alerts | $49 a month | Per seat, not published, ask | Free | Not published, ask |
How the five were judged
A sourcing tool is only as good as what it reads and what it can prove. Six questions separate them, and each row of the table above answers one.
- Is it built for the profile you hire? A tool that covers every job covers researchers too, thinly. Specialisation shows in the filters: first authorship, venue, subject, stage of the doctorate.
- Does it read the work, or the self-description? Most researchers never write a profile. Their record is the papers they sign, and those are public.
- Does it know one person from two? The same researcher signs "J. Chen", "Jing Chen" and "Jing X. Chen" on three papers. A tool that does not merge them shows three weak profiles instead of one strong one.
- Can you check every claim? A link to the paper, the repository or the thesis turns a ranking into something a hiring manager can verify in a minute.
- Does it know when they become available? A researcher contacted in the second year of a doctorate says no. The same person contacted six to nine months before the defence reads the message.
- What does it cost per useful name? A free tool that takes a day per shortlist is not free.
1. Founding Hires, for AI researchers
The Founding Hires index is built for one search: AI researchers who are about to come on the market. It is ours, so read the next paragraphs as a claim to check against the method page and the atlas.
It reads the published literature, not profiles, and merges every name variant into one researcher: 1 027 515 people on 554 206 papers. Each profile links to the papers on the subject you searched, the venue, the author position and the university on the latest paper. For 19 781 researchers a public source dates the end of the doctorate, and every date carries its evidence: a filed thesis, a statement on a personal page, or an estimate from academic age, labelled as such. Saved searches become alerts that write to you when a date is confirmed or a defence moves under six months.
| 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
It covers the applied side of AI too. Applied AI engineers and AI engineers sign papers as middle authors, publish models and datasets, and leave research for industry on a date the index can read. Where it fits less well: software engineers with no AI record at all. They are outside the literature by definition, and that is what StarHunt is for.
2. StarHunt, for software engineers
StarHunt applies the same principle to code. It reads GitHub in real time and ranks software engineers on what they commit: the stack, the languages, recent activity and the projects other people depend on. It is the tool for the engineers who build products, back ends and infrastructure, and who never appear on a paper.
The pool is large and mostly passive. StarHunt's State of Developer Reachability 2026, measured across 275,000 public GitHub profiles, found that only 14% of developers signal they are open to work, while 61% of the active ones publish an email address or a LinkedIn profile anyone can use. StarHunt is also ours. For an AI hire, researcher or engineer, the Founding Hires index remains the better tool.
Hiring software engineers?
Search GitHub by what people build.
StarHunt ranks software engineers on their commits, their stack and their recent activity, and shows how to reach them.
3. LinkedIn Recruiter
LinkedIn Recruiter is the generalist. It covers every profession in every market, and its strength is reach: most working engineers and many industry researchers keep a profile, and InMail gets a message in front of them.
Its limit for AI research is that it reads what people write about themselves. Doctoral students rarely maintain a profile, a paper appears only if the member lists it, and nothing on the profile says when a doctorate ends. It fits a company hiring across many roles, with AI one of them. It fits less well when the candidate is known for a paper and not for a job title. Pricing is per seat and not published.
4. Google Scholar
Google Scholar is free and indexes almost every paper on the web. For reading one researcher's record, their papers, their co-authors and how often they are cited, it is the fastest tool there is.
It is not built for sourcing. Author profiles exist only for researchers who created one, so many strong early-career names stay unlinked strings on a paper. There is no filter by country, stage or availability, no contact detail beyond a verified email domain, and no official export for a list. Use it to check a shortlist, not to build one.
5. SeekOut
SeekOut is a talent search platform for technical recruiting teams. It aggregates public professional profiles, including technical signals such as code, and adds the search and outreach tools a large recruiting team needs.
It fits a team running many technical searches at once, where breadth and workflow matter more than depth on one field. It fits less well when the question is who in a narrow research subject defends in the next nine months, which needs the literature and a dated end of doctorate. Pricing is not published; ask.
Which tool for which hire?
- A researcher on a precise subject. Search the literature, rank on first authorship, and contact people six to nine months before the defence. The Founding Hires index is built for this, and Google Scholar helps check the finalists.
- An AI engineer or an applied AI engineer. Look for middle authorships, published models and a recent move into industry. The Founding Hires index.
- A software engineer. Read the code. StarHunt.
- Many roles, AI among them. Reach and workflow matter most. LinkedIn Recruiter or SeekOut.
- A mixed team. AI profiles from the index, software engineers from GitHub, and a generalist tool for everything else.
Not sure which profile the problem calls for? Start with AI researcher, applied AI engineer or AI product engineer. Would rather hand the search over? See the best AI recruitment agencies, compared.
Common questions
- What is the best tool to find AI researchers?
- A tool that reads the published research rather than self-written profiles, merges each researcher's papers into one profile, links every claim to its source and dates the end of the doctorate. The Founding Hires index does all four. Google Scholar is free and useful to check one researcher. LinkedIn Recruiter covers every profession but relies on what members write about themselves.
- Can I find AI researchers on LinkedIn?
- Some. Industry researchers often keep a profile. Doctoral students and postdocs rarely do, their papers appear only if they list them, and no profile says when a doctorate ends. For early-career researchers, search the literature instead.
- Is Google Scholar enough to source AI researchers?
- It is enough to read one researcher's record, not to build a list. Profiles exist only for researchers who created one, and there is no filter by country, stage or availability, and no export.
- What is the best tool to find AI engineers?
- One that reads the work behind the title. Many AI engineers sign papers as middle authors or publish models, and the Founding Hires index finds them there, one profile per person, with the end of the doctorate when there is one. For software engineers without an AI record, StarHunt reads GitHub and ranks them on their stack and recent activity.
Search the index
Every AI researcher, one profile each, with a date.
Search by subject, see every paper behind a name, and get an alert when a defence comes closer.