Find the exact people you need, from inside your own logged-in LinkedIn
Describe who you need. Your AI searches live people surfaces in your browser and returns a shortlist with a why and a source for each.
“Find staff-level ML engineers in fintech who are open to advising. Give me a why for each.”
R. Almeida
Staff ML Engineer · payments scale-up
Why: leads fraud-model team, lists “advising” in their headline.
J. Tran
Staff ML Engineer · lending startup
Why: shipped a risk model, posts publicly about MLOps.
M. Köhler
Staff ML Engineer · neobank
Why: 8 yrs fintech ML, “open to advisory roles” in About.
From one brief to a sourced shortlist
Connect the AI you already use, ChatGPT, Claude, or Gemini, to your browser. From one plain-language brief it runs four steps and ends at a sourced shortlist.
- 1
Describe who you need in plain language
Skip boolean strings and saved-search filters. Say it like a brief, such as “staff ML engineers in fintech who are open to advising,” and your AI turns it into the searches it needs to run.
- 2
It searches the live people surfaces, not a cached index
It runs LinkedIn people search with your filters, then widens to conference speaker pages, GitHub profiles, personal sites, and podcast guest lists, wherever the right people actually show up.
- 3
It reads each profile and attaches a reason
For every candidate it opens the live page and pulls the detail that proves the match, like a current role, a recent talk, or public writing, so the shortlist comes back with a WHY, not just a name.
Keyword databases match strings. This matches meaning.
- 4
You get a sourced shortlist of people you didn’t know existed
Names, roles, and a one-line reason land as a clean list, each with a source link you can open. Discovery ends here. Enrichment, outreach, and email lookup live on sibling pages.
Your AI wants to export your shortlist (8 people) to your sheet.
Export shortlist to your sheet
8 people · name · role · why · profile link
Each row carries a source URL · nothing is contacted
The shortlist is yours, and you approve where it goes
Discovery ends at a verified list, and it never messages anyone. When you’re ready to push the shortlist into your own sheet or tool, that write happens like every consequential Actionbook action: you see exactly what it will export before it commits.
- Shortlist, not outreach: The stopping point is a sourced list of people. No connection request or email is ever sent.
- Approve the export: Pushing the list to a sheet surfaces for your approval first, in plain language.
- Hand off cleanly: Send approved people to enrichment, prospecting, or email lookup as separate, on-purpose steps.
Built for whoever has to find the people
One workflow: describe who you need, get a sourced shortlist back.
Technical recruiters
Source past the same database everyone queries. Describe the candidate by skill, stack, and signal, and get live profiles, including people no keyword search surfaces.
Leads fraud-model team, lists “advising.”
Shipped a risk model, posts on MLOps.
8 yrs fintech ML, “open to advisory.”
Founders hiring first roles
No sourcing team yet. Say who you need for role one, and your AI returns a shortlist of real people with a reason each, so the first conversations start from a warm list.
Journalists finding sources
Find experts who actually publish on a niche, not the names off a PR list. The agent reads bylines, talks, and public posts, and hands back people you can quote with a link to why.
Discovery, not a database lookup
It reads the live surfaces where people actually appear, like LinkedIn search, speaker pages, GitHub, and personal sites, surfacing net-new people no months-old index would return.
Event & podcast organizers
Book speakers and guests by what they’ve covered, not who you already follow. It mines speaker pages and guest lists for fresh names that fit your topic.
Prompts to start finding people
Copy a prompt, paste it into your AI, and Actionbook runs it in your own browser session.
Find 10 staff-level ML engineers in fintech who look open to advising. Search LinkedIn people search with my filters, open each profile, and return a shortlist with a one-line why and a profile link for each.
Find speakers who covered AI agents at recent conferences. Read the speaker pages, pull their name, the talk, and where they work now, and give me a sourced list with a reason each is relevant.
Find maintainers of popular Rust crates who write publicly about their work. Check their GitHub profiles and personal sites, and return 8 people with a link and a one-line why for each.
Shortlist candidates matching this job description from live profiles, not a stale database. Run LinkedIn people search, read each match, and return name, current role, and a why-they-fit line with a link.
Find potential podcast guests who have spoken about RAG and retrieval on other shows. Mine guest lists and personal sites, and return a shortlist of fresh names with a source link and a reason each.
Find researchers and practitioners who write publicly about payments fraud. Read their profiles and recent posts, and give me 8 people I could interview, each with a why and a link to their work.
Frequently asked questions about AI people search
What people ask before letting an AI find net-new people for them.