KiHire.
AI recruiting for healthcare — jobs, candidates, matches. First market: physiotherapy.
Recruiting in healthcare is broken.
Physiotherapy practices, outpatient centers, and clinics have roles open all year round and no in-house recruiter. LinkedIn Recruiter costs 9,000 € a year per seat — a volume that doesn’t justify the spend. External headhunters take 25 % of annual salary. Neither option matches the reality of small providers.
The result: roles stay open for months, applications fall through the cracks, hires happen through personal networks — with all the bias and bottlenecks that come with that.
Agents that search, read, decide — not just filter a list.
KiHire takes a job spec, lets a Claude agent extract the context (skills, seniority, industry), launches a sourcing agent that finds candidates, and scores each one with a match rationale. Outreach mails are personalised based on candidate experience.
Beta running in the physiotherapy first market. Per-match pricing, no seat model — fair for providers hiring 1–10 roles a year. The data layer runs on a dedicated Supabase backend, fully migrated from the Notion beta.
/ tech-decisions
Three decisions that mattered.
Agent Loop
Claude sub-agents per sourcing step
Instead of one long prompt chain: specialised agents for job analysis, profile search, match scoring, outreach drafting. Each agent has its own context window — better results on long job specs.
Data Layer
Supabase as the data backbone
Fully migrated from the Notion beta to Supabase — around 15,000 job postings on file, deterministic matching. Postgres + Auth + Edge Functions, one project per app.
Pricing
Per match, not per seat
Classical ATS tools bill per recruiter — punishes small companies with irregular hiring needs. KiHire bills per qualified match. Fair model for 1–10 hires per year.
/ other apps
Cookmark
Recipes from Instagram reels, turned into structured cards — in one click.
Watchlist
Streaming PWA with a filter stack: what we watch tonight, without 20 minutes of scrolling.
GOLAZO
Tournament platform for my own youth team at the Gothia Cup — skills, duels, World-Cup predictions. Live with 16 real users.