JobJitsu

JobJitsu Logo

Inspiration:

We noticed how intimidating and inconsistent career-fair and interview prep tools are for students. We wanted to create something that helps students practice real conversations — not just answer flashcard questions.

What it does:

JobJitsu simulates realistic recruiter chats, generating tailored questions and follow-ups using AI. It gives instant feedback, tracks progress, and helps users improve communication confidence.

How we built it:

We used FastAPI for the backend, MongoDB for session storage, and React for the front end. Gemini powers question generation, ElevenLabs provides natural recruiter voices, and Supabase Auth handles user authentication.

Challenges we ran into:

Parsing AI-generated JSON reliably, coordinating front-end and back-end timing with async audio generation, and managing session data consistency were tough. Making the recruiter feel human was also a challenge.

Accomplishments that we're proud of:

We built a full end-to-end pipeline — users can start a session, have a recruiter conversation, and receive feedback — all in under 24 hours. The voice interaction and adaptive follow-up system really brought it to life.

What we learned:

We learned how to sanitize and validate AI responses, use multiple APIs together efficiently, and manage rapid teamwork across database, backend, and front-end components.

What's next for JobJitsu:

Expanding into behavioral and technical interview modes, adding analytics dashboards for improvement tracking, and deploying a public beta for university students.