Automated Job Hunting via Gmail & AI

Most people still manage their job search manually. They update spreadsheets, search through old emails, and constantly reconstruct the current status of each application.

Gmail is already the source of truth. Instead of asking people to enter the same information again, the product should turn incoming emails into editable job cards.

Job Tracker converts job-related emails into a visual workspace. Each application becomes a card that can be reviewed, edited, and moved through the hiring process.

By scanning the email inbox with the Gmail API. GPT-4O-MINI reads, understands, and extracts the relevant job information, then generates an editable card. By keeping the original email connected to each card, users can always return to the context behind an application.

I built and tested Job Tracker with real inbox data, iterated on the Gmail scanning flow, and presented the working beta at Index Chinatown.

Classification results are cached in the database, allowing future scans to skip previously processed emails and reduce both processing time and token usage.