If you cannot audit what the vendor measures, you cannot ship it. AGPL core you can self-host — same product as Cloud.
Security and works-council reviews fail for the same reason: you cannot show what the vendor measures, where people data goes, or how to run the stack without handing them the keys. A closed HR cloud that also claims to score AI usage is a non-starter.
AGPL core you can self-host and audit, the same product surface as Cloud, and a public page for what measurement never touches. With Coverage, the skill file for a step is a file you download and run in your own tools: the platform executes nothing and sees no agent usage. The modules below are the talent system on that foundation — skills and assessments you can inspect, not endpoint monitoring.
The same platform modules as on the landing — only the ones that act on the risk above.
Describe a process. See which steps an agent can take, which stay with people, and where nobody covers it yet.
Build living competency frameworks and see what your workforce can actually do.
Self, 180° and 360° assessments with calibration and detailed reporting.
Knowledge tests with generated questions and automatic scoring.
Hire for demonstrated skills — resume scoring, skill checks, interview analysis. The recruiter decides.
Framework generation, semantic search and suggestions across every module.
A sandbox preloaded with a sample team, its assessments, open vacancies and two processes already broken down into steps — no installs, nothing to connect. Your own data comes later, on your terms.
