People decide
Our AI recommends, drafts and explains. Decisions that significantly affect people — hiring, lending, health — stay with people, and anyone can ask for a human review.
We say when it is AI
Our agents introduce themselves as AI on calls, in chats and in meetings. Disclosure is on by default in every product.
Privacy by design
We collect the least data needed, keep it for the shortest time, and never use customer data to train shared models.
Evaluated before and after release
Every AI feature has evaluations for accuracy, safety and cost that must pass before release, and it is monitored in production for drift.
Fairness
Where AI assesses people, we remove irrelevant signals, test outcomes across groups and share the results with the teams who use it.
Security
We defend against prompt injection and data leakage, encrypt data, restrict access and review our systems regularly.
Raising a concern
If you believe one of our AI systems has behaved unfairly, unsafely or inaccurately, write to singh.rajendar2023@gmail.com. We investigate every report.