QA Testing an Internal AI Assistant

Context

An internal AI assistant was being evaluated for its ability to provide accurate answers using internal organizational knowledge.

My Role

Used domain expertise to test responses, identify inaccuracies, and communicate issues to product and TPM stakeholders.

Example

During testing, identified a compliance-sensitive knowledge/ownership error that could have directed users toward incorrect internal information and escalated it through the appropriate product/TPM channel.

Outcomes

• —Surfaced a compliance-sensitive accuracy issue before broader rollout
• —Contributed structured, domain-grounded feedback to the product/TPM team