Interview Question Generator: Technical Whitepaper

Version: 1.2.0
Status: In Production (with known deployment issue)
Last Updated: January 16, 2026

Executive Summary

The Interview Question Generator is a web-based tool that generates behavioral interview questions mapped to GrowthFlowEngineering's proprietary GFE Skill System. It enables hiring managers and recruiters to build structured interview guides based on specific competency tasks across Sales, Marketing, and Finance domains.

Key Metrics

188 tasks covered across 3 domains
346 interview questions with scoring rubrics
5 skill levels (L0-L4: Apprentice to Partner)
API response time: <200ms average

1. Problem Statement

The Interview Quality Gap

Most organizations face three critical problems in technical/behavioral interviewing:

Generic Questions: Interviewers ask the same vague questions ("Tell me about yourself") regardless of role
No Calibration: Different interviewers assess candidates using different criteria
Missing Rubrics: No standardized "what good looks like" guidance for scoring responses

Our Solution

The Interview Question Generator addresses these problems by:

Mapping questions to specific competency tasks (e.g., "S-230: CRM Hygiene & Deduplication")
Providing "what good looks like" criteria for each question
Including follow-up probes to dig deeper into candidate responses
Organizing by skill level so questions match the seniority being assessed

2. System Architecture

2.1 Technology Stack

Component	Technology	Purpose
Frontend	VitePress + Vanilla JS	Static site with client-side interactivity
Backend	Netlify Functions	Serverless API endpoints
Data Store	JavaScript object (embedded)	Question bank compiled into function
Source of Truth	TSV file	Master question bank for editing

2.2 Data Flow

┌─────────────────────────────────────────────────────────────┐
│                    GFE-SkillSystem Repo                      │
│  specs/interview-questions/interview-questions.tsv          │
│  (Master question bank - 346 questions)                      │
└─────────────────────────────────────────────────────────────┘
                              │
                              │ generate-function.cjs
                              ▼
┌─────────────────────────────────────────────────────────────┐
│                growth-flow-engineering Repo                  │
│  netlify/functions/interview-questions.js                   │
│  (Compiled function with TASK_QUESTIONS object)             │
└─────────────────────────────────────────────────────────────┘
                              │
                              │ Netlify Deploy
                              ▼
┌─────────────────────────────────────────────────────────────┐
│                    Production API                            │
│  /.netlify/functions/interview-questions                    │
│  Actions: list, generate, all                               │
└─────────────────────────────────────────────────────────────┘
                              │
                              │ fetch()
                              ▼
┌─────────────────────────────────────────────────────────────┐
│                    Frontend Page                             │
│  /tools/interview-questions                                 │
│  VitePress page with DOM manipulation                       │
└─────────────────────────────────────────────────────────────┘

2.3 File Structure

GFE-SkillSystem/
└── specs/
    └── interview-questions/
        ├── interview-questions.tsv    # Master question bank (source of truth)
        ├── README.md                   # Documentation
        ├── QUESTION_IMPROVEMENT_GUIDE.md  # Quality framework
        ├── generate-function.cjs      # TSV → Netlify function generator
        └── improve-questions.cjs      # Batch quality improvement script

growth-flow-engineering/
├── netlify/
│   └── functions/
│       └── interview-questions.js     # Serverless API (3274 lines)
└── docs/
    └── en/
        └── tools/
            ├── index.md               # Tools landing page
            └── interview-questions.md # Frontend UI (667 lines)

3. API Reference

Base URL

https://growthflowengineering.xyz/.netlify/functions/interview-questions

3.1 List All Tasks

Returns all available tasks for building the selection UI.

Request:

bash

GET /.netlify/functions/interview-questions?action=list

Response:

json

{
  "success": true,
  "action": "list",
  "totalTasks": 188,
  "levels": {
    "0": { "name": "Apprentice", "focus": "Learning fundamentals..." },
    "1": { "name": "Practitioner", "focus": "Executing playbooks..." },
    "2": { "name": "Architect", "focus": "Designing systems..." },
    "3": { "name": "Strategist", "focus": "Setting direction..." },
    "4": { "name": "Partner", "focus": "Shaping organization..." }
  },
  "tasks": [
    {
      "id": "S-001",
      "taskId": "S-001",
      "title": "Qualify Inbound Leads",
      "taskTitle": "Qualify Inbound Leads",
      "domain": "sales",
      "level": 1,
      "level_name": "Practitioner",
      "levelName": "Practitioner",
      "question_count": 2,
      "questionCount": 2
    }
    // ... 187 more tasks
  ]
}

3.2 Generate Questions

Returns interview questions for selected tasks.

Request:

bash

POST /.netlify/functions/interview-questions
Content-Type: application/json

{
  "action": "generate",
  "taskIds": ["S-001", "S-230", "M-100"],
  "questionsPerTask": 2
}

Response:

json

{
  "success": true,
  "action": "generate",
  "totalTasks": 3,
  "totalQuestions": 6,
  "results": [
    {
      "taskId": "S-001",
      "taskTitle": "Qualify Inbound Leads",
      "domain": "sales",
      "level": 1,
      "levelName": "Practitioner",
      "questions": [
        {
          "question": "Describe your lead qualification framework...",
          "whatGoodLooksLike": "References BANT/MEDDIC/similar framework...",
          "probes": ["How do you handle borderline leads?", "What's your MQL to SQL conversion rate?"]
        }
      ]
    }
  ]
}

3.3 Get All Data

Returns complete question bank (useful for exports/debugging).

Request:

bash

GET /.netlify/functions/interview-questions?action=all

4. Question Quality Framework

4.1 The Problem with Template Questions

Initially, we auto-generated questions for tasks without hand-crafted content. The template approach produced generic, useless questions:

BAD (Template):

"How do you set strategy for CRM hygiene & deduplication?" What good looks like: "Clear strategic thinking, balances multiple priorities, aligns with org goals"

This question is worthless because:

No specific scenario
No numbers or constraints
Generic rubric that could apply to anything

4.2 Quality Improvement Framework

We developed a 5-point framework for high-quality questions:

Principle	Description	Example
SPECIFICITY	Names real tools, metrics, scenarios	"Your CRM has 50,000 contacts but a 15% bounce rate..."
BEHAVIORAL ANCHOR	Uses STAR format	"Tell me about a time you found duplicate records causing..."
TENSION/TRADE-OFF	Forces judgment	"Sales wants Gong, marketing wants Chorus, finance says pick cheapest..."
MEASURABLE OUTPUT	Concrete assessment criteria	"Prioritizes by impact, uses waterfall enrichment, sets auto-suppression..."
LEVEL-APPROPRIATE	L0 ≠ L3 question complexity	Apprentice: "How do you log activities?" vs Strategist: "Design the hygiene strategy"

4.3 Before/After Examples

S-230: CRM Hygiene & Deduplication

Before	After
"How do you set strategy for CRM hygiene?"	"Your CRM has 50,000 contacts but a 15% email bounce rate. Walk me through how you would clean this database in the next 30 days."
Generic rubric	"Prioritizes by impact (high-value accounts first), uses waterfall enrichment (ZoomInfo → Clearbit → manual), sets bounce threshold for auto-suppression, creates ongoing hygiene cadence not one-time fix"

F-132: Deal Profitability & Margin Analysis

Before	After
"How do you approach deal profitability?"	"Sales is pushing to close a $500K deal at 15% discount plus free implementation. Walk me through your margin analysis framework."
Generic rubric	"Calculates net margin after all costs (COGS, implementation, CS, payment terms), models LTV vs CAC, assesses precedent risk, proposes alternative structures"

5. Current Issue: Frontend Not Loading Tasks

5.1 Symptom

The frontend shows "Loading tasks..." indefinitely, even though the API returns 188 tasks correctly.

5.2 Root Cause Analysis

Timeline of changes:

Original API returned task.id, task.title, task.level_name (snake_case)
New API returned task.taskId, task.taskTitle, task.levelName (camelCase)
Frontend expected the old format, couldn't find properties, failed silently

Debugging steps taken:

✅ Verified API returns 188 tasks with correct data
✅ Fixed API to return BOTH formats (dual properties: id AND taskId)
✅ Fixed frontend to accept either format: task.id || task.taskId
⏳ Waiting for Netlify to rebuild and deploy the VitePress site

5.3 The Fix (Applied)

API Fix (interview-questions.js):

javascript

function getAllTasks() {
  return Object.entries(TASK_QUESTIONS).map(([id, task]) => ({
    // Dual properties for backward/forward compatibility
    id: id,
    taskId: id,
    title: task.title,
    taskTitle: task.title,
    level_name: LEVELS[task.level]?.name || 'Unknown',
    levelName: LEVELS[task.level]?.name || 'Unknown',
    question_count: task.questions.length,
    questionCount: task.questions.length,
    domain: task.domain,
    level: task.level
  }));
}

Frontend Fix (interview-questions.md):

javascript

taskGrid.innerHTML = filtered.map(task => {
  const taskId = task.id || task.taskId;
  const taskTitle = task.title || task.taskTitle;
  const levelName = task.level_name || task.levelName;
  const questionCount = task.question_count || task.questionCount;
  // ... rest of render
}).join('');

5.4 Verification Steps

API is working (confirmed via curl):

bash

curl "https://growthflowengineering.xyz/.netlify/functions/interview-questions?action=list" | grep totalTasks
# Returns: "totalTasks":188

Dual properties are present (confirmed):

bash

curl ... | grep '"id":'
# Returns both "id" and "taskId" for each task

Frontend build pending: Netlify must rebuild the VitePress site for the frontend fix to deploy.

6. TSV Data Format

6.1 Schema

Column	Type	Description
task_id	string	Unique identifier (e.g., "S-230")
task_title	string	Human-readable task name
domain	enum	"sales", "marketing", or "finance"
level	int	0-4 (Apprentice to Partner)
question_number	int	1 or 2 (multiple questions per task)
question	string	The interview question text
what_good_looks_like	string	Assessment criteria
probe_1	string	First follow-up question
probe_2	string	Second follow-up question

6.2 Example Row

tsv

S-230	CRM Hygiene & Deduplication	sales	3	1	Your CRM has 50,000 contacts but a 15% email bounce rate...	Prioritizes by impact (high-value accounts first)...	How do you balance speed vs accuracy?	What downstream systems would be affected?

7. Development Workflow

7.1 Adding/Editing Questions

Edit the TSV (source of truth):

bash

code GFE-SkillSystem/specs/interview-questions/interview-questions.tsv

Regenerate the function:

bash

cd GFE-SkillSystem/specs/interview-questions
node generate-function.cjs

Commit both repos:

bash

# GFE-SkillSystem (TSV changes)
git add specs/interview-questions/
git commit -m "feat: update interview questions for [task-ids]"
git push

# growth-flow-engineering (function changes)
git add netlify/functions/interview-questions.js
git commit -m "feat: regenerate interview questions from TSV"
git push

7.2 Quality Improvement Batch

To improve multiple template questions at once:

Add improved questions to improve-questions.cjs
Run the script:
bash
```
node improve-questions.cjs
```
Regenerate and commit as above

8. Future Enhancements

8.1 Planned Features

Feature	Priority	Status
Export to PDF	High	Not started
Save interview guides	Medium	Not started
AI-assisted question generation	Medium	Not started
Candidate self-prep mode	Low	Not started
Integration with ATS	Low	Not started

8.2 Question Bank Expansion

Current coverage: 188/188 tasks (100%)
Hand-crafted questions: ~50 tasks
Template questions: ~138 tasks (need improvement)
Target: 100% hand-crafted by Q2 2026

9. Conclusion

The Interview Question Generator transforms how GrowthFlowEngineering clients conduct competency-based interviews. By mapping questions to the GFE Skill System's 188 tasks, providing concrete assessment criteria, and organizing by skill level, we enable consistent, high-quality candidate evaluation.

Current Status

✅ API fully functional (188 tasks, 346 questions)
✅ Backend fixes deployed
⏳ Frontend fix awaiting Netlify build completion
📋 Ongoing: Improve template questions with strategy consulting-quality content

Contact

For questions or contributions, contact: Support@growthflowengineering.xyz

This whitepaper is auto-updated as the Interview Question Generator evolves.

Interview Question Generator: Technical Whitepaper ​

Executive Summary ​

Key Metrics ​

1. Problem Statement ​

The Interview Quality Gap ​

Our Solution ​

2. System Architecture ​

2.1 Technology Stack ​

2.2 Data Flow ​

2.3 File Structure ​

3. API Reference ​

Base URL ​

3.1 List All Tasks ​

3.2 Generate Questions ​

3.3 Get All Data ​

4. Question Quality Framework ​

4.1 The Problem with Template Questions ​

4.2 Quality Improvement Framework ​

4.3 Before/After Examples ​

5. Current Issue: Frontend Not Loading Tasks ​

5.1 Symptom ​

5.2 Root Cause Analysis ​

5.3 The Fix (Applied) ​

5.4 Verification Steps ​

6. TSV Data Format ​

6.1 Schema ​

6.2 Example Row ​

7. Development Workflow ​

7.1 Adding/Editing Questions ​

7.2 Quality Improvement Batch ​

8. Future Enhancements ​

8.1 Planned Features ​

8.2 Question Bank Expansion ​

9. Conclusion ​

Current Status ​

Contact ​

Interview Question Generator: Technical Whitepaper

Executive Summary

Key Metrics

1. Problem Statement

The Interview Quality Gap

Our Solution

2. System Architecture

2.1 Technology Stack

2.2 Data Flow

2.3 File Structure

3. API Reference

Base URL

3.1 List All Tasks

3.2 Generate Questions

3.3 Get All Data

4. Question Quality Framework

4.1 The Problem with Template Questions

4.2 Quality Improvement Framework

4.3 Before/After Examples

5. Current Issue: Frontend Not Loading Tasks

5.1 Symptom

5.2 Root Cause Analysis

5.3 The Fix (Applied)

5.4 Verification Steps

6. TSV Data Format

6.1 Schema

6.2 Example Row

7. Development Workflow

7.1 Adding/Editing Questions

7.2 Quality Improvement Batch

8. Future Enhancements

8.1 Planned Features

8.2 Question Bank Expansion

9. Conclusion

Current Status

Contact