I developed TrackIt to bring job applications, recruiter contacts and interview preparation into one place. My work covers the React frontend, with a particular focus on the FastAPI backend, data and AI integration.

AI that supports the job search

An assistant that understands the context

The assistant uses application and contact information to provide responses tailored to the job search.

• Choice of model: Run Ollama locally, or use OpenAI, Gemini and Claude with a personal API key.
• LangChain integration: connecting AI to relevant application information.
• Two languages: conversations in French or English.

Interview practice tailored to the role

Practice built around the selected role, with feedback to help improve from one session to the next.

• Feedback on every answer during practice.
• Saved sessions to revisit previous practice.
• PDF report highlighting strengths and areas for improvement.

One workspace for the whole search

Track job applications

A dashboard to review statuses, save favourites, search and filter opportunities, and identify applications that need a follow-up.

Save a job listing from Chrome

The Chrome extension saves job listings from compatible websites so they can be found in TrackIt.

Keep contacts and documents together

Recruiter contact details and attachments stay linked to applications, making it easier to prepare conversations with the right information.

Export your data

Applications can be exported to a structured Excel file, alongside the PDF reports from interview practice.

My implementation: backend, data and AI integration

REST API and authentication

• FastAPI: development of the REST API behind the application’s features.
• JWT and HttpOnly cookies: authentication implementation.
• React: frontend design and API integration.

Storage and AI orchestration

• SQLite: data storage for the local application.
• PostgreSQL: database for the Docker environment.
• LangChain: AI integration using application and contact context.
• Ollama or a personal API key: a choice between local execution and an external provider.

From API to distributable application. TrackIt is available on macOS, Windows and Linux with a local web interface and integrated updates. I automated testing and cross-platform builds with GitHub Actions. The downloadable version requires no prior installation of Docker, Python or Node.js; a Docker environment is also provided for developers.

Open-source project: the code is available on GitHub and version 1.0.0 is available to download.

Download TrackIt ↗ View code on GitHub ↗