Built for the rules on employment and AI
Employment decisions are high-risk by law. Orbio is designed for that from the first line of code: humans decide, agents inherit permissions, every action is logged, and the evidence your security and legal teams need is self-serve.
Agents act. Your people decide.
A person makes every employment decision
Agents interview, summarise and recommend. Hiring, rejecting or ending a relationship is done by a named person, on the record. There are no automated rejections.
People know it’s an agent
Disclosed up front, in the candidate’s or employee’s language. A human is available on request at any point in the conversation.
Every recommendation shows its source
Each conclusion links to the conversation it came from. Your team can read the transcript before acting.
Independently audited for bias
Screening outcomes are monitored by group with adverse-impact analysis every quarter and an independent annual audit (NYC LL 144 compliant).
Your data never trains models
Customer data is used to run your deployment and nothing else. No cross-customer training, no vendor model training.
Human gates are configurable
Which actions need approval, by whom, at what threshold — set per entity, site and role.
Employment and AI rules, by jurisdiction. Tracked continuously; documentation available in the Trust Center.
Your security review, answered before the first call.
Reports, policies, sub-processors, DPA and questionnaire responses — self-serve at trust.orbio.work.