Moderated user interview

Interview

Help a neighboring product team conduct interviews for "Personal Account AI Gateway", a web app for internal AI agent developers
The service was meant to track agent implementation status, provide training materials, and enable work with DevOps
The product's key goal was to reduce time to market and the number of support tickets

Defined the research goal — to identify the current pain points teams face during the AI agent rollout process
Prepared a research guide: expected outcome, decisions the team could make based on the results, research questions, hypotheses to test, research method, respondent requirements, and the list of interview questions
Aligned the guide with the team and got contacts for suitable respondents
Conducted the research and, based on the insights gathered, formulated conclusions and proposals for improving problem areas

Method: in-depth interviews, 40–60 minutes, remote, with video and audio recording
Respondent criteria: priority given to those who had already gone through the rollout process, or those preparing to do so and familiar with the documentation
Conducted 8 interviews with POs and analysts responsible for the AI agent rollout process

Developers don't fully understand the structure and requirements described in the existing documentation
When facing difficulties, developers turn to colleagues for help more often than to the documentation
Developers don't understand what stage their agent is at (submitted, reviewed, approved, etc.)
The lack of a unified interface makes it harder to track progress and fragments the rollout process

Conducted 8 interviews with team members (POs, analysts) responsible for the rollout process
Confirmed 3 out of 4 hypotheses
Identified key pain points in the rollout process and interface wishes
Proposed solutions to the identified problems
Presented the research results to the team