| name | user-interview-analysis |
| description | Analyze user interview transcripts from clinical trial stakeholders (CROs, sponsors, sites) to extract themes, pain points, opportunities, and persona insights. |
| allowed-tools | mcp__claude_ai_Atlassian__createConfluencePage, mcp__claude_ai_Atlassian__searchConfluenceUsingCql |
| argument-hint | [Paste transcript text, or provide path to transcript file] |
User Interview Analysis
Analyze user interview transcripts from clinical trial EDC stakeholders to produce structured, actionable insights for the Talosix product team.
Workflow
-
Receive Input: The user provides one or more interview transcripts (pasted text or file paths). Read the content carefully.
-
Identify the Persona: Classify the interviewee into one or more Talosix persona categories:
- Site Coordinator / CRC: Enters data, manages visits, resolves queries
- Principal Investigator (PI): Reviews and signs off on data, medical decisions
- Clinical Data Manager (CDM): Designs CRFs, manages data cleaning, locks databases
- Clinical Monitor / CRA: Performs SDV, conducts monitoring visits
- Sponsor Project Manager: Oversees trial execution, tracks enrollment/timelines
- CRO Operations Lead: Manages multi-study operations, resource allocation
- Regulatory/Quality: Ensures compliance, manages validation documentation
- Biostatistician: Consumes clean datasets, defines derivations
- IT/System Administrator: Manages system configuration, integrations, user access
-
Analyze and Generate the structured output below.
Output Structure
Interview Summary
- Interviewee: Role, organization type (sponsor/CRO/site), therapeutic area
- Interview Date: [date if available]
- Interviewer: [name if available]
- Duration: [if available]
- Key Context: Study phase, study size, current EDC system (if mentioned)
Themes Identified
Group findings into themes. For each theme:
- Theme Name: Brief label (e.g., "Query Resolution Burden", "eCRF Navigation Friction")
- Description: 1-2 sentence summary
- Supporting Quotes: Direct quotes from the transcript (with attribution)
- Frequency/Intensity: How prominently this theme appeared
- Relevance to Talosix: How this maps to current or planned features
Pain Points
For each pain point:
- Pain Point: Clear description
- Severity: Critical / High / Medium / Low
- Frequency: Daily / Weekly / Per-study / Occasional
- Current Workaround: What does the user do today?
- Quote: Verbatim supporting quote
- Product Implication: What Talosix feature or improvement could address this
Opportunities
- Opportunity: Description of unmet need or improvement area
- User Value: What would change for the user
- Business Value: Revenue, retention, competitive differentiation
- Feasibility Estimate: Quick win / Medium effort / Large initiative
Workflow Observations
- Document the user's current workflow step-by-step
- Note friction points, inefficiencies, and manual steps
- Identify where Talosix EDC fits (or could fit) in their workflow
Competitive Intelligence
Extract any mentions of:
- Other EDC systems (Medidata Rave, Veeva Vault EDC, Oracle InForm, Castor, REDCap)
- Specific likes/dislikes about competitor products
- Switching triggers or barriers
Recommendations
- Immediate Actions: Quick wins the product team should pursue
- Feature Requests: Specific features or improvements suggested
- Further Research: Areas that need deeper investigation
- Persona Updates: New insights about this persona's needs, goals, and frustrations
Multi-Interview Synthesis
When analyzing multiple transcripts together, also produce:
Cross-Interview Theme Matrix
| Theme | Interview 1 | Interview 2 | Interview 3 | Prevalence |
|---|
| [Theme] | Mentioned/Not | Mentioned/Not | Mentioned/Not | X/N |
Persona Refinement
- Behavioral patterns observed across interviews
- Segment differences (sponsor vs. CRO vs. site perspective)
- Priority ranking of needs by persona
Actionable Summary
- Top 3 product priorities based on the research
- Confidence level (strong signal / emerging signal / weak signal)
- Recommended next steps (more research, prototype, build)
Tips
- Preserve exact quotes -- do not paraphrase when citing evidence
- Flag contradictions between interviewees
- Note emotional intensity (frustration, excitement, indifference)
- Distinguish between stated needs and observed/implied needs
- Pay attention to clinical trial-specific regulatory pain points (audit trail fatigue, validation burden, 21 CFR Part 11 friction)
