All projects
Built & TestedAutomationvisibility: sanitizedIndependent project
Automated Job Search Application
Streamlit app that aggregates, filters and de-duplicates job postings
Locally working application; not deployed publicly.
- Role
- Developer
- Links
- Source is private or sanitized; no public repository is linked.
- Reading mode
Project overview
A local Streamlit application that queries the JSearch API through RapidAPI for multiple role titles and locations, scores results for relevance against configurable keywords, removes duplicates that appear across queries and sources, and presents a clean, sortable list. API keys are loaded from environment variables and never committed.
Problem and context
Searching several job boards daily for overlapping role titles produces noise and duplicates; automating collection and filtering saves time and surfaces better matches.
Objectives
- Query multiple titles and locations in one run
- Score relevance with configurable keyword weights
- De-duplicate across sources
- Keep secrets out of the codebase
My contribution
- Designed query batching and rate-limit handling
- Implemented relevance scoring and de-duplication
- Built the Streamlit UI and export
Technology stack
- Python
- Streamlit
- JSearch API
- RapidAPI
- requests
- pandas
- python-dotenv
Results
- Working local application that produces a de-duplicated, ranked job list
Limitations
- Depends on third-party API availability and quotas
- Not deployed publicly