Personal software
Career Autopilot: Internship Search & Ranking
Built a personal internship-search app that gathers listings across employer sites and job feeds, then ranks them around my engineering background and career goals.
Personal application · Internship discovery and prioritization
My contribution
I built a local dashboard around the way I search for internships: mechanical engineering roles, adjacent technical fields, my target term, and practical location preferences. The work connects source collection, duplicate detection, adjustable ranking, scheduled checks, and a browser extension for reviewing application fields.
Results & observations
The captured dashboard contains 3,655 internship listings, alongside scholarship and event tracking. I can review ranked leads, new listings, and approaching deadlines from the same screen. That count is a snapshot of the collection; availability, eligibility, and fit still need to be checked against the original posting.
Fewer tabs. More useful leads.
An internship search spreads quickly across employer boards, public lists, job feeds, and alert emails. I wanted one place to collect those leads and decide what deserved a closer look, without repeating the same search on every platform.
Collect, reconcile, rank
Source adapters feed a common opportunity model. Duplicate detection compares job identifiers, normalized URLs, and employer/title matches while preserving differences between recruiting terms. Configurable ranking weighs fit, pay, career trajectory, deadlines, and regional cost of living; the interface exposes the reasons behind each score.
A workflow around the search
The Python and FastAPI application stores its working collection in SQLite. Scheduled checks find new listings, quiet hours hold notifications overnight, and reminders flag approaching deadlines. A companion browser extension supports user-reviewed form filling, with employer-specific behavior still being tested.
Checking listings and rankings
I use the rankings to choose which roles to investigate, then check availability and requirements on the employer’s site. Incomplete listings and changing source formats need ongoing attention as I refine the collection.
Tools & methods
- Python
- FastAPI
- SQLite
- API integration
- Data deduplication
- Weighted ranking
- Browser extensions
- Task scheduling
- Interface design
