Meet Hunter, the tireless head of research for your firm. He reads the brief, hunts every source you give him in parallel, verifies every candidate against two or three independent sources, and hands your team a tiered shortlist ready for your CRM.
A senior research agent for recruitment firms, built on a disciplined sourcing methodology and run in production by our team. Hunter does the mechanical search and verify work. Your researchers do the judgement.
No line about a candidate without a URL to back it. Unsourceable facts are flagged or omitted.
Live since March 2026, running real roles, with 24+ failure modes already solved and codified.
We build, secure, run, and tune Hunter week over week. Your team just reviews the shortlists.
A full day of sourcing compressed to about an hour of agent work and 25 minutes of review.
Hunter is the opposite. He is already in production, running real roles, with a source on every claim.
A competent researcher burns 8 to 20 hours per role on discovery and verification before any judgement work begins. The mechanical search and stitch is most of that, and it compounds across every role.
Researcher labour on discovery and verification alone, before any judgement work begins.
Same brief, two researchers, different longlists. Invisible until it shows up downstream as wasted outreach.
LinkedIn, a verified email, and the firm site all disagree. Reconciling takes time; skipping it produces stale longlists.
LinkedIn, Brave, Exa, Apollo, Lusha, your ATS, and whatever else you run. The switching cost is paid every role.
You review and decide. Hunter does the searching and stitching.
Send it by Telegram, paste it, or pull it straight from your CRM. I read it end to end, extract the underlying role, and flag inconsistencies before I search.
Title variants, adjacent backgrounds, sector synonyms, firm-size band, trajectory signals. You confirm or correct me in chat. This is the first quality gate, and errors caught here are cheap.
Structured filters, semantic match, known-target firm sweeps, sector mapping. Six sources and counting in a typical setup, and the list grows with your stack. You end up with an 80 to 150 candidate longlist, each with a current role, firm, location, and source URL.
Each shortlisted candidate confirmed against two independent sources within 60 days. Tier 1 gets a verified work email and phone from a third source. I surface discrepancies, I never bury them.
Tier 1 is the top 3 to 5, fully verified, each with a specific standout. If there are fewer than three genuine Tier 1s, I say so. I never promote a Tier 2 to fill the slot.
Delivered in your CRM's schema, with a top-10 preview dropped in chat and a full audit trail filed. Every candidate traceable to the search that surfaced them.
A full day of researcher work, compressed to about an hour of Hunter's work and 25 minutes of your team's review. You review and decide. Hunter does the searching and stitching.
These are Hunter's standards. He enforces them on himself, every role, whether or not anyone is checking.
Hunter confirms every shortlisted candidate against two independent public sources within 60 days. No exceptions.
Hunter won't write a single line about a candidate without a URL to back it. Unsourceable facts are flagged or omitted.
No fabricated candidates, roles, or histories. He won't stitch fragments from different people into one. Uncertainty is surfaced, never hidden.
Hunter never promotes a candidate to fill a slot. Fewer Tier 1s with strong evidence beats more with thin evidence.
These aren't marketing claims. They're enforced at the tool layer. Hunter literally cannot write a candidate to a CSV without a source URL.
Most firms stall at Tooled. Hunter moves your whole research function up the curve.
Researchers source by hand across a few tabs. Every role starts from zero. Quality rides on who's free that week.
A LinkedIn seat, a scraper, maybe an LLM writing Booleans. Faster in spots, but nothing reconciles sources or holds a standard.
Hunter runs the mechanical sourcing end to end, verified and tiered, with an audit trail. Researchers move to judgement and client work.
Hunter's methodology is tuned to your firm and improves every week. Throughput climbs, the standard holds, and your best researchers place more.
A real role Hunter ran, anonymised at the client's request.
"This is the v3. I'd have spent 6 hours getting here last year. Sending it to the client tomorrow."
2-week setup: dedicated infrastructure, your data providers wired in, and your methodology, tier rules, and voice codified into Hunter. Live shadowing on 2–3 real roles before cutover.
Hunter, running, tuned, and supported. New methodology rules codified live, no downtime. Throughput and quality reviewed monthly.
You pay the data providers (Brave, Exa, Apollo, Lusha, and any others you add) directly per credit. Hunter reports every credit spent. No markup.
You own your role data, your methodology document, and your workspace state. Stop anytime and take everything.
No. Hunter absorbs the mechanical search and verify load, 60 to 80% of sourcing time, so your researchers spend their hours on judgement, candidate engagement, and client work. You hire researchers for the parts that need a human. Hunter isn't that.
Verified. Hunter confirms every shortlisted candidate against two independent sources within 60 days, with a URL on every claim. Tier 1 candidates get a verified work email and phone from a third source. He surfaces discrepancies rather than producing a clean-looking but stale list.
He surfaces the conflict instead of guessing. If LinkedIn, a verified email, and a firm site disagree, Hunter flags it, shows you what each source says, and won't assert a fact he can't back. Uncertainty is visible, never hidden.
Atlas/Recruit integrations are established (read-only). Bullhorn, Loxo, JobAdder, Vincere, or custom CRMs are a per-integration build, typically 3 to 5 days. If you'd rather not integrate, Hunter delivers CSV in any schema you need.
Every claim is sourced, so you can check anything in seconds. Beyond that, we maintain a live regression log: 24+ real failure modes already diagnosed and permanently prevented at the tool layer. Hunter gets more reliable the longer he runs, not less.
Hunter doesn't contact candidates, make hiring decisions, assess cultural fit, or move candidates between CRM stages. He produces verified, tiered, source-cited candidate data and hands it to your team. Outreach and judgement stay with your humans.
Book a discovery call. Walk us through a real role, and we'll show you exactly what Hunter would surface, verify, and deliver, and how many researcher hours he gives back.