Interview Coach
AI-powered dual-mode interview simulator with terminal CLI and voice-enabled local Web Dashboard.
Practice mock interviews from your terminal or browser.
A developer-focused, privacy-first tool supporting Groq, OpenAI, Anthropic, and offline Ollama. Practice using text or voice, track progress, and export professional evaluation reports.
Featured Project
Interview Coach
Project Resources
Explore live deployments, documentation, and source code.
CLI & Web Dashboard
Platform
Stable (v2.0.0)
Stage
Role Agnostic
Interview Types
Automated & PDF
Reports
Problem
Interview preparation is often inconsistent, expensive, and privacy-invasive.
Most mock interview tools lock you into cloud subscriptions, upload personal interview transcripts, or require heavy AI models and voice libraries to run locally.
Privacy Concerns & Lock-in
Candidates are forced to upload responses to third-party services and rely on fixed, paid AI subscriptions.
Fragile Local Setups
Adding voice support offline traditionally requires installing complex system binaries like FFmpeg or downloading huge ML models.
Lack of Structured Feedback
Standard chatbots lack repeatable, guided workflows that grade performance, spot missed topics, and suggest ideal responses.
Capabilities
Structured, voice-enabled interview simulations with offline LLMs.
Dual-Mode Architecture
Switch between a blazingly fast text CLI in your terminal and a local web dashboard in your browser.
Browser-Native Voice
Voice practice powered by standard browser Web Speech APIs (STT & TTS) with zero heavy python dependencies or external API costs.
Bring Your Own Model
Practice with Groq, OpenAI, Anthropic, offline Ollama, or any custom OpenAI-compatible endpoint.
Answer Evaluation & Insights
Get detailed breakdowns with scores (0–100), verdicts, strengths list, omitted topics, and customized study guides.
Privacy & Local Persistence
Configuration, interview logs, and local analytics databases reside 100% on your local machine with zero telemetry.
Professional PDF Export
Compile structured evaluation feedback, charts, and ideal answers into a print-ready, clean PDF report.
Install
Install globally via npm.
Run realistic interview simulations directly from your terminal using npm or npx.
Technology
Built with
Architecture
Unified speech and assessment engine.
A local pipeline that hooks into your preferred LLM provider, manages stateful conversation history, and evaluates responses.
Configuration Layer
Interactive terminal prompt or dashboard panel to set active model providers, API credentials, and generation timeouts.
Dual-Mode Interface
CLI rendering text menus via Inquirer, or a local Express server powering a multi-tab web application.
Speech & Interaction
Captures speech using webkitSpeechRecognition and plays questions aloud via SpeechSynthesisUtterance.
Evaluation Pipeline
Scores answers against target skills, stores history in local JSON databases, and builds printable PDF report files.
Evolution
Project evolution.
Terminal Prototype (v1.0.0)
Built a basic command-line simulation using Groq SDK, basic session storage, and text-only role practice.
Local Evaluation Engine
Developed the scoring system, correct/incorrect verdicts, missed discussion topics, and ideal study guide answers.
Dual-Mode & Express Server (v2.0.0)
Upgraded to a local Express server backend, transitioning voice inputs to browser speech APIs for ultimate reliability.
PDF Reports & Offline support
Implemented client-side PDF document compiles, visual charting libraries, and custom timeouts for local Ollama weights.