AI Technical Interview Agent 🎯

An adaptive, accessible AI Technical Interview Agent designed for candidates completing the 31-Day Enterprise AI Engineering Cohort.

---

🌟 Overview

Preparing for enterprise AI engineering interviews requires more than memorizing static Q&A. Candidates must demonstrate deep architectural understanding, explain trade-offs, solve live coding challenges, and defend design decisions made across modern AI disciplines.

The AI Technical Interview Agent simulates a real technical interview with a Lead AI Architect:

• Personalized Candidate Context: Analyzes cohort candidate profiles (role, experience, completed missions, skipped days, attempt counts) to start interviews at candidate-specific entry points.
• Adaptive Deep-Dives: Asks a minimum of 8 questions covering at least 4 distinct curriculum days, triggering dynamic follow-up questions when responses lack architectural detail.
• Live Code Testing Tool: Dynamic coding challenges triggered during the interview with interactive problem statement, timer, and solution evaluator.
• Google AI Studio BYOK Architecture: Powered by Google AI Studio (default model gemma-4-31b-it) with optional model switching (gemini-2.5-flash, gemini-2.5-pro) and BYOK API key configuration.
• Ground-Up Accessibility: Fully WCAG 2.1 AA compliant with high-contrast modes, font scaling, speech dictation (STT), voice playback (TTS), complete keyboard shortcuts, and screen reader announcements.
• Actionable Evaluation Report: Generates structured performance reports with domain radar mastery scores, readiness levels, engineering strengths, growth areas, and targeted curriculum study roadmaps.

---

🚀 Features

• 20 Cohort Candidate Profiles: Pre-loaded candidate data across senior engineers, junior developers, DevOps specialists, and business analysts.
• 31-Day Curriculum Mapping: 8 modules covering Vector Databases, RAG, Prompt Engineering, Agentic AI, MCP Protocol, Guardrails, and Docker/Kubernetes Deployment.
• Live Code Testing Tool: Interactive live code editor with countdown timer and automated test evaluation.
• Accessibility Suite:
  • Alt + S : Toggle Voice Dictation (Microphone Input)
  • Alt + R : Read Current Question Aloud (Text-to-Speech)
  • Alt + C : High Contrast Theme Toggle
  • Alt + F : Finish & View Feedback Report
  • Esc : Close Modals / Navigation
• Export & Print: Export formatted PDF evaluation report with domain mastery charts and transcript breakdowns.

---

⚙️ Quick Start

1. Install Dependencies

pip install fastapi uvicorn pydantic httpx

2. Run Backend Server

python app.py

3. Open Portal in Browser

Navigate to http://localhost:8000 to launch the interactive interview portal.

---

📡 API Specification

Endpoint	Method	Description
/api/health	GET	Health check endpoint
/api/candidates	GET	Returns list of cohort candidates
/api/curriculum	GET	Returns 31-day curriculum data
/api/interview/start	POST	Starts a new candidate interview session
/api/interview/chat	POST	Processes candidate message & returns adaptive AI response
/api/interview/live-test/submit	POST	Evaluates submitted live coding solution
/api/interview/end	POST	Concludes interview & returns feedback report
/api/interview/session/{id}/feedback	GET	Retrieves formatted evaluation report

---

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.