Job interviews are stressful enough. Preparing for them shouldn't require a subscription, uploading your data to someone else's servers, or being locked into one AI provider.
That's why I built Interview Coach, an open-source AI interview simulator you can run with any AI provider you trust or fully offline using local models.
It adapts to whatever role you're interviewing for: engineering, product, design, marketing, finance, and more.
Why I Built It
Most AI interview tools I tried shared the same problems:
- Locked to one AI provider
- Little or no support for local models
- Limited customization
- Weak privacy story
- Not built with developers in mind
I wanted something you could install, configure once, and use with whichever AI provider you preferred.
What It Does
Interview Coach comes as both a terminal tool and a local web dashboard. It supports:
- Cloud AI providers or local models via Ollama
- Voice-based interviews
- Session history that resumes where you left off
- Performance reports with PDF export
Everything is built to be simple, fast, and privacy-first.
Hands-Free Voice Interviews
The web dashboard uses your browser's built-in speech features — no extra installs, no platform-specific setup. Open it and start talking.
Bring Your Own AI Model
Use a cloud API or run everything locally with Ollama. Handy if you want to try different models, keep your interview data private, cut API costs, or prep with no internet at all.
Practice Any Role
It's not just for software engineers. Interview Coach generates role-specific interviews for product management, design, marketing, HR, finance, operations, customer success, and custom roles — adjusting questions based on the role, experience level, focus areas, and how the interview's been going so far.
Feedback That Actually Explains Itself
Every answer gets a score, a verdict, what you did well, what you missed, and what a stronger answer would look like — not just "correct" or "incorrect."
Everything Saved Locally
Progress is saved automatically as local files. Close the app mid-interview and pick up right where you stopped. Nothing gets uploaded anywhere.
Reports You Can Actually Use
After each session: a performance report, question-by-question scores, a skills breakdown, and a PDF export — so you can track improvement over time.
What's Next
Some things on the roadmap: multi-language interviews, resume-aware interview generation, company-specific interview modes, coding interview support, team dashboards, and richer report visualisations.
Try It / Contribute
Interview Coach is fully open source — feedback and contributions welcome.
- GitHub: https://github.com/J3rry320/interview-coach
- npm: https://www.npmjs.com/package/interview-coach
Got ideas for new interview modes or features? Let me know in the comments.
Interview Coach is an open-source product from Code Media Labs, our software and media design studio. We do everything from business applications and SaaS platforms to Media Shoots and AI-powered workflows, and this project reflects our approach to creating thoughtful, production-ready software.
And as always, Happy Coding!
Top comments (0)