Applied AI · Research interest
Reliable Enterprise AI Systems
How AI agents keep organizational context, and where a person still has to approve the work.
Abstract
An interest in the reliability of AI agents inside enterprise software: context, evaluation, permissions, and human oversight. This is not an active study.
Research questions
- How do coding agents keep a consistent picture of an organization's systems?
- What kind of human approval is actually useful in a production workflow?
Publication status
Research interest. No manuscript.