Backend · Utrecht
Backend · Berlin
You are paying a few hundred euro a month for a stack that hands you 200 profiles — and then you spend the evening throwing most of them away. Leadl.recruit hands back a scored shortlist with the reasoning written out, ready to forward to your client.
5 free searches a month · no card · no credits · no annual contract
Every candidate comes back with a 0–100 fit score built from real work — code, technical answers, project history — not from a self-written profile. The bottom of the list is already gone before you open it.
Each score carries its reasoning: which requirements are confirmed, which are unconfirmed, what the red flags are and what to ask. It reads as an artifact of how you selected — not as a number you have to defend.
One run is one role, including up to ten deep dossiers. Contact details are part of a dossier, not a separate balance — so a hard search costs exactly what an easy one costs. From €49 a month.
The question behind every client call
On contingency you are not paid for the search — you are paid for the placement. Which means a client who quietly doubts your shortlist is a client who keeps two other agencies on the same role.
You cannot answer that doubt with a longer list. You answer it by showing how the selection was made: what was searched, how many people were scored against the brief, which requirements each shortlisted candidate actually evidences, and why the near-misses did not make it.
Leadl.recruit produces that as a by-product of doing the work. Every shortlist you send can carry the reasoning behind it — so the conversation moves from “trust me” to “here is the selection, look at it yourself”.
Illustrative example of a candidate dossier. Yours is generated from your own vacancy.
The stack maths
A boutique desk ends up with a seat for search, a seat for the local job market, something for the AI layer, and ChatGPT to glue it together — a few hundred euro a month before a single fee lands.
And the final step still lands on you: opening profile after profile to decide which of them are actually worth a message. That step is the product here, not a feature bolted onto a database.
50 searches a month, one seat, cancel at the end of any month. The free plan gives five searches with no card, which is enough to test it on a role you are working right now.
How a run works
No boolean strings to maintain, no credit budget to ration, no dashboard to learn on a Sunday.
The real one your client sent you, in whatever shape it arrived. The requirements become the scoring criteria.
GitHub, StackOverflow, Discord and Reddit communities and the open web — public professional activity, not only profile headlines.
0–100 on skills evidence, experience relevance, requirement coverage and red flags — each with its reasoning attached.
Shortlist with interview questions per candidate. Outreach from Solo up, ATS export from Team. The verdict stays yours.
It does not decide for you, and it does not run interviews. It automates the first pass — the sorting hours — and leaves the judgement where your client is actually paying for it: with you. Community sourcing is a complement to GitHub, StackOverflow and LinkedIn, not a replacement: not every strong candidate is publicly active, and the score says when the evidence is thin instead of guessing.
Roles outside the English-speaking bubble
Tools built around English-language profile keywords go thin the moment the role is in Dutch, German or Polish, or the specialism is narrow enough that the right twenty people never wrote a keyword-optimised headline in their lives.
Scoring from published work — repositories, technical answers, project history, public community activity — keeps working where headline matching stops, because the evidence is in what someone built, not in how they described it.
A profile headline is a marketing document. A merged pull request, an answered thread and a shipped project are not.
Which is why the fit score is built from the second kind of evidence — and tells you plainly when a candidate has none of it.
Pricing for a 1–5 person desk
One run is one search for one role, including up to ten deep dossiers. Contact discovery is included, monthly billing, cancel at the end of any month.
The part of the stack you use to build a long list and then re-filter it by hand. Leadl.recruit sources across GitHub, StackOverflow, Discord, Reddit and the open web, scores every candidate against your vacancy with the reasoning attached, and returns a shortlist. Most boutique desks keep one seat somewhere for messaging and drop the rest.
Neither. You pay per search. Contact discovery is part of a dossier rather than a separate balance, so a hard role costs the same as an easy one and you are never rationing credits mid-search. See the credit-model comparison.
That is what it is built for. Every candidate carries a 0–100 fit score with its reasoning: confirmed requirements, unconfirmed claims, red flags and the questions worth asking. It is a selection artifact, not an opaque match percentage.
It ranks people on what they publish rather than on English job-title keywords in a headline, which is exactly where LinkedIn-first tools go thin. It remains a complement to your own market knowledge, not a substitute for it.
You paste a vacancy and get a scored shortlist back in minutes, instead of spending the evening opening profiles. The first pass is automated; the verdict is not.
You are told, and offered a run pack or pay-as-you-go at €1.50 a run. Nothing is charged automatically unless you switch that on.
Paste the live vacancy, see the scored shortlist and the reasoning behind it. Five searches free, no card, no call with a sales rep.