Skip to main content

AI Interviewer

The AI Interviewer is Celeredge's conversational evidence-collection feature. Some of the most important inputs in consulting aren't written down — they live in stakeholders' heads. The AI Interviewer captures that by holding a topic-based conversation with a stakeholder and turning their answers into grounded, citable evidence the engagement can use.

It's an AI feature in the same family as the rest of the intelligence layer: what it captures becomes evidence that assessments, signals, and Ask the Data retrieve and cite — see Citations & Trust.

Who can access​

Members of the engagement team launch interviews. Interviews are designed to be completed by stakeholders, including people outside the core team.

Inputs → processing → outputs​

  • Inputs: the topics you want covered, and a stakeholder to answer.
  • Processing: the AI conducts a conversational interview — asking, following up, and steering toward the topics that matter for the assessment — and extracts the substance from the conversation.
  • Outputs: structured evidence attached to the engagement, retrievable and citable like a document passage. It can directly fill the gaps an assessment surfaces.

Where you launch it​

  • Diagnose — as part of the evidence step before running an assessment (section 3 of the Diagnose checklist).
  • Data & Evidence → Collect data — topic-based collection at any time.

Step by step​

  1. From Diagnose or Data & Evidence → Collect data, start an AI interview.
  2. Choose the topics to cover.
  3. Share it with the stakeholder to complete conversationally.
  4. As responses come in, they become evidence on the engagement.
  5. Run or re-run your assessment so the new evidence is included.

Expected behavior & limitations​

  • The interviewer focuses on the topics you set — give it the right scope to get useful evidence.
  • Evidence from an interview only affects assessment results on the next run — re-run after the interview completes.
  • Like all Celeredge evidence, interview output is attributable: answers trace back to the interview as their source.

Best practices​

  • Use it to close evidence gaps. When an assessment returns Not grounded answers, an interview on that topic is often the fastest way to fill the gap.
  • Target the right person. Direct each interview at the stakeholder who actually holds the knowledge.
  • Keep topics focused. Tighter topics produce sharper, more citable evidence.