We don't spray CVs. We use a candidate knowledge graph and LLM matching to find the precise fit — then place them.
The graph finds what manual recruiters miss. Humans close what the graph finds.
Every candidate conversation is parsed into the knowledge graph — 200+ signals covering skills, stack, salary, work style, and career trajectory.
When a role arrives, the model queries the graph and scores fit across every signal — not just keywords. The output: one to three precise matches.
We brief the candidate, align on salary, and manage the process to offer. Relationships and negotiation stay human.
| Meridi | Traditional recruiters | Job boards | |
|---|---|---|---|
| Shortlist | 1–3 candidatesmatched & verified | 5–10 CVs | Hundreds of applicants |
| Matching | Knowledge graph + LLM200+ signals per candidate | Keyword search + memory | Keyword filters |
| First shortlist | 48 hours | 1–2 weeks | You do the screening |
| Fee | 15% · contingency only | 20–25%, often retained | Subscriptions + ad spend |
| Guarantee | 3-month replacement | Varies by contract | None |
30 minutes on the role: stack, seniority, salary, team context.
The model scores every candidate in the graph against the brief.
1–3 candidates, verified, briefed, and salary-aligned.
We coordinate scheduling, feedback, and negotiation.
Backed by a 3-month replacement guarantee.