← Back
Hire Me Pls
Description
Hire me pls helps remove some of the cognitive load of job applications while simultaneously allowing candidates to stand out. We employ context about job applicants and context about the company/role they are applying to in order to generate a perfectly customized cover letter in just seconds.
Note: The linked version of the web application is currently deactivated because of API call cost constraints but check out the working demo from the hackathon here
Role
I had the idea for the project and recruited three other developers at PennApps to assist me in building out the project.
Skills Utilized
- Google Cloud Run
- Docker
- Github
- Python
- FastAPI
- GPT-4 API
- Metaphor API
- Prospeo API (for linkedin profile scraping)
- TailwindCSS
- Netlify
Contribution
Had vision for final product and coached team to reach our goal assigning tasks to each individual which aligned with their strengths. Build backend of application with FastAPI and containerized the application to deploy to GCP. Prompt engineered to combine Prospeo API and Metaphor API in a GPT-4 api call to generate cover letter. Streamed response from GPT back to frontend for better UX than waiting for entire cover letter to be generated (can take up to a minute).
Outcome
Won "best use of metaphor API", check out the Devpost. Had daily active users after the hackathon completed (had to shut down because of API calls exceeded free tiers). Metaphor shared the project on their Linkedin and X
Key Takeaways
Scoping of your project is crucial especially in hackathons. Have a bias for getting things working (don't optimize too early). Companies like linkedin take in your data but then won't provide a way for you to get it back out... 🫠. They ban scraping and I had to go through third party APIs just to get info from someone's profile page which is so uncool.
Further Information
I am continuing the development of this project under a new name Employable AI. Feel free to checkout the Github as well to monitor progress on this project