Our AI analyses missions and profiles to surface the relevant matches. It explains its recommendations, but it never decides in the recruiter's place.
The published mission is structured into comparable criteria: must-have and nice-to-have skills, technologies, seniority and years of experience, location and work mode (onsite, hybrid, remote), languages, contract type (freelance or permanent), budget or salary range, and start date.
Each profile is compared to the mission criterion by criterion, each criterion carrying a weight defined in advance. Must-have skills weigh the most; the other criteria (technologies, experience, seniority, location and work mode, languages, rate, contract type, availability) refine the result. The calculation always follows the same rules: with the same data, the outcome is always the same.
The reasons shown come straight from that calculation: only criteria that are actually met become reasons, technical ones first, in a few readable lines. A language model may refine the ordering of a shortlist from already-anonymised data; when it is unavailable, the deterministic ranking is kept.
The recruiter reviews, compares, discards or requests a contact. TalentAI never contacts, accepts or rejects a talent automatically: the AI recommends, the human decides.
The AI recommends. The human decides.
Hosting, encryption, payments and GDPR are detailed on a dedicated page.