Inspiration
HR tech tools often need realistic employee data and resumes, but real data is private — and fake data looks fake. FakeHR solves this with AI: generating synthetic employee data and clean, professional resumes using Google Gemini.
What it does
-> Log in securely -> Generate synthetic employee data (customizable) -> Generate one-page resumes with Google Gemini -> Download in CSV, JSON, Excel, or ZIP (PDFs) ->Preview resumes before download -> Leave feedback and view history
How we built it
-> Frontend: Streamlit -> AI Integration: Google Gemini 1.5 Flash -> PDF Resumes: Jinja2 + pdfkit + wkhtmltopdf -> Database: SQLite -> Security: bcrypt password hashing -> Deployment: AWS EC2 (24×7 live)
Challenges we ran into
-> Gemini API rate limits → fallback logic with Faker -> Conversion of HTML file into pdf for resume generation -> Deployment on AWS
Accomplishments that we're proud of
-> Successful integration of Google Gemini for clean, structured JSON resume generation -> A smart fallback system using Faker when the Gemini API limit is exceeded -> Multi-format downloads including PDFs, CSV, Excel, and ZIP files -> A feedback system with rating storage and session history using SQLite
What we learned
-> Streamlit optimization -> AWS deployment -> Resilient design under API limits
What's next for FakeHR
-> Global data support (Canada, UK, India) -> API access for integration -> More smooth and fast results -> Make it more user intearctive
Built With
- amazon-web-services
- api
- bcrypt
- faker
- gemini
- pdfkit
- python
- sqlite
- streamlit
- wkhtmltopdf
Log in or sign up for Devpost to join the conversation.