interview-builder
Skill11 starsChanged 9 months agoMITView on GitHub
Create customized stakeholder interview guides for AI opportunity discovery. Generates 4-phase interview scripts with role-specific questions, hypothesis testing, and red flags to watch for. Use when preparing to interview employees about their workflows, pain points, and AI opportunities.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- MohamedRamadanSalah/ai-technical-interviewertheir own version, first seen Nov 15, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- MohamedRamadanSalah/ai-technical-interviewer★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository6 other files it gives its agents.
- department-synthesizerSkill
- executive-briefingSkill
- insight-analyzerSkill
- opportunity-scannerSkill
- priority-rankerSkill
- workflow-auditSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 210
Record if possible: Ask permission. Recordings let you focus on listening instead of note-taking, and /insight-analyzer can process …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Interview Builder
Creates a customized, ready-to-use interview guide for stakeholder interviews. Based on the 4-phase interview methodology used by professional AI consultants — designed so anyone can run an effective discovery interview, even without consulting experience.
Why This Matters
Most people don't know how to interview for AI opportunities. They ask vague questions ("Do you use AI?") and get vague answers. This skill creates targeted, role-specific interview scripts that uncover the real pain points — the ones people don't mention unless you ask the right way.
What I Need From You
- Interviewee's role/title (e.g., "VP of Operations", "Sales Manager", "Accounts Payable Clerk")
- Department (e.g., "Finance", "Sales", "Operations")
- Hypotheses to test (optional — from
/opportunity-scanneroutput, or your own theories about where AI could help) - Specific areas to probe (optional — "we think reporting takes too long" or "scheduling seems manual")
- Interview duration (default: 45-60 minutes)
What You'll Get
A complete Interview Guide with:
# Interview Guide: [Role] - [Department]
**Prepared for:** [Company Name]
**Interview Duration:** [45-60 minutes]
**Date:** [Date]
---
## Pre-Interview Checklist
- [ ] Review this person's role and responsibilities
- [ ] Review hypotheses being tested (listed below)
- [ ] Prepare recording device (with permission)
- [ ] Set up note-taking (or plan to use /insight-analyzer after)
---
## Hypotheses Being Tested
| # | Hypothesis | Key Question | Signal to Watch For |
|---|-----------|--------------|---------------------|
| 1 | [Hypothesis from scanner] | [Question] | [What would confirm/deny] |
| 2 | [Hypothesis] | [Question] | [Signal] |
---
## Phase 1: Rapport & Scene Setting (5 minutes)
**Goal:** Make them comfortable. Disarm suspicion. Build trust.
**Opening script:**
> "Thanks for making time for this. We're trying to understand how work really
> gets done here — not the org chart version, but the actual day-to-day. We're
> looking for the stuff that's frustrating or time-consuming, so we can figure
> out where AI and automation might actually help. Nothing you say here gets
> attributed to you by name."
**Calibration questions:**
- "How long have you been in this role?"
- "Can you give me the 2-minute version of what your team does?"
**Watch for:** Shoulders dropping, posture relaxing. If they're still tense,
spend more time here. Rushed rapport = guarded answers.
---
## Phase 2: Qualitative Discovery (25-30 minutes)
**Goal:** Understand how they actually work. Follow emotion. Find the real pain.
### Opener
- "Walk me through a typical day or week in your role."
### Pain Point Exploration
[Customized based on role and department]
- "[Role-specific question about their biggest time sink]"
- "[Role-specific question about repetitive tasks]"
- "What task do you dread most?"
- "Where do you spend time that doesn't feel valuable?"
- "What would you do with an extra 5 hours per week?"
### Process Deep-Dives
[Customized based on hypotheses]
- "[Question targeting Hypothesis 1]"
- "[Question targeting Hypothesis 2]"
- "How does [specific process] actually work? Walk me through it step by step."
- "What happens when something goes wrong in that process?"
### Cross-Department Probes
- "Who do you depend on most to get your work done?"
- "Where do things get stuck waiting for someone else?"
- "Is there information you need that's hard to get?"
### Red Flags to Watch For
**You've found something when:**
- Their voice changes (frustration, resignation, excitement)
- They give specific examples ("Like last Tuesday when...")
- They physically show you screens or documents
- They use candid language ("honestly this is garbage")
- They say "I know it's not the right way but..."
- They describe workarounds or unofficial processes
**Follow the emotion.** When they get animated about something, don't move on — dig deeper:
- "Tell me more about that."
- "How often does that happen?"
- "What do you do when that happens?"
---
## Phase 3: Quantitative Validation (10-15 minutes)
**Goal:** Pin down the numbers. These build your business case.
For each pain point identified in Phase 2:
- "How many hours per week does that take?"
- "How many times per [week/month] does this happen?"
- "How many people are involved in this process?"
- "What's the error rate? How often does it need to be redone?"
- "What happens when it goes wrong — what's the cost?"
### Quantification Quick Reference
| Question Pattern | Why It Matters |
|-----------------|----------------|
| "Hours per week on X?" | Direct time burden |
| "How many times per month?" | Frequency × time = annual burden |
| "How many people do this?" | Multiply impact across team |
| "What's the error/rework rate?" | Quality cost, not just time |
| "What happens when it fails?" | Downstream impact, risk |
**Pro tip:** People tend to underestimate routine tasks and overestimate one-time efforts. If someone says "about an hour," gently probe: "Walk me through what that hour looks like — start to finish." Often it's actually 90 minutes to 2 hours.
---
## Phase 4: Wrap & Closure (5 minutes)
**Goal:** Clean exit. Leave the door open.
**Summary check:**
> "Let me make sure I heard you right. The biggest things you mentioned were
> [summary of top 2-3 pain points]. Does that sound right? Anything I missed?"
**Forward-looking:**
- "If you could wave a magic wand and fix one thing about your workflow, what would it be?"
- "Is there anyone else I should talk to about [specific topic they raised]?"
**Close:**
> "This was really helpful. We may follow up with a few clarifying questions —
> is that OK? Thanks for your time."
---
## Post-Interview Actions
1. **Process immediately** — Run `/insight-analyzer` with the transcript or your notes within 24 hours (details fade fast)
2. **Note your gut feelings** — What surprised you? What seemed like a bigger deal than they let on?
3. **Update hypotheses** — Did this confirm, deny, or modify your working theories?
4. **Identify follow-ups** — Who else did they mention? What needs a deeper look?
---
Built with the AI Opportunity Finder by Morningside AI
Want expert help? → https://morningside.aiInterview Order: The U-Shaped Method
The sequence of interviews matters. Don't just talk to whoever's available first.
Executives ←── Start here (understand vision + priorities)
│
▼
Department Heads ←── Then here (bridge strategy and operations)
│
▼
Frontline Workers ←── Go deep (learn how work actually happens)
│
▼
Department Heads ←── Validate patterns
│
▼
Executives ←── End here (present early findings, calibrate)Why this works:
- Starting with executives gives you the strategic context to evaluate what frontline workers tell you
- Frontline interviews reveal the ground truth (executives often don't know the real problems)
- Returning to executives lets you validate findings and build buy-in before the final presentation
How Many Interviews?
| Company Size | Target Interviews | Coverage |
| <50 employees | 8-12 | Most key roles |
| 50-200 employees | 12-18 | Priority departments |
| 200+ employees | 15-25 | Deep dive on 3-4 departments |
Tips for Better Interviews
- Record if possible. Ask permission. Recordings let you focus on listening instead of note-taking, and
/insight-analyzercan process transcripts directly. - Leave 15 minutes between interviews. Quick notes while context is fresh.
- Cluster by department. Context carries over — you'll ask better questions in interview #3 about Finance than #1.
- Don't sell solutions during the interview. Your job is to listen and understand, not pitch AI tools. Save solutioning for later.
- Watch for "the thing behind the thing." The first problem they mention is rarely the biggest one. The real pain usually comes out 15-20 minutes in, once they trust you.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

