Know why a candidate fits. Not just that they do.

A score for every candidate against the real requirements of the role, plus the reasoning behind it: which requirements are met, which are partly met and what is missing. You get a ranked shortlist in minutes and a written case for each person that you can send to a hiring manager or client as it is.

Yes. That is the point. Each score comes with plain-language reasoning tied to the job requirements, so you can show a client exactly why a candidate is on the shortlist and where you still want to verify something. It reads like a recruiter's recommendation, not a black-box number.

No. Pickr matches on a professional skills graph, so related skills and different vocabulary for the same capability are recognized. A candidate who has done the neighboring job is surfaced with the reasoning for why they might fit, and verified qualifications are shown separately from claims on a CV. You still make the call.

Yes. Every interview, rejection and placement teaches Pickr what a strong candidate looks like for that role and that client, and the next shortlist reflects it. The reasoning stays visible throughout, so an improved ranking is still one you can explain to the people who asked for it.

Pickr is hosted in Frankfurt, Germany, and names and emails are redacted from AI prompts by default. Candidate matching is treated as a high-risk AI system under the EU AI Act: candidates are informed before it is used, every score comes with reasoning and no hiring decision is ever fully automated. A data processing agreement (AVV) is included.

Yes. Connect Claude or ChatGPT to Pickr via MCP and you can search candidates, move them through the pipeline, schedule interviews and send emails from a chat window, without opening the app. It runs on the subscription you already have.