README.md

metadata

title: RecSys Skills
emoji: 📚
colorFrom: yellow
colorTo: blue
sdk: gradio
sdk_version: 5.49.1
app_file: app.py
pinned: false
license: mit

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference SkillSense - Graph-based Skill Recommendation System

Challenge Tackled: Structured skill profile generation with evidence and confidence scores ✅ Recommendation of skills based on users who share similar interests (this enables skill-gap analysis) ✅ Limited training data: user interactions, interested roles, skills stored in a graph to improve future recommendations✅ Mutli-source support- system takes pdf, docx format of resume, professional summaries, portfolio or linked pages as pdf, docx ✅ Scalability- graph expands as more users interact with the system, improving the quality of recommendations ✅

Setup & Deployment (Developer Instructions)

The application can be run locally or deployed on Hugging Face Spaces. Below are the steps for both workflows.

1. Clone the Repository git clone cd

2. Create and Activate a Virtual Environment (Recommended) python3 -m venv venv source venv/bin/activate # Mac/Linux venv\Scripts\activate # Windows

3. Install Dependencies pip install -r requirements.txt

4. Configure Secrets

The application uses environment variables for API keys (Neo4j credentials).

Local Development:

Create a .env file:

NEO4J_URI= NEO4J_USER= NEO4J_PASSWORD=

Hugging Face Deployment:

On your Space page:

Settings → Variables and Secrets → Add Secrets

Add each key exactly as referenced in app.py.

5. Run Locally python app.py

The UI will open in your browser (typically at http://127.0.0.1:7860).

Deploying to Hugging Face Spaces

Go to: https://huggingface.co/spaces

Click Create New Space

Choose:

Space SDK: Gradio

Visibility: Public or Private

Either:

Upload your project files, or

Select “Repository → Add Files → Upload directory”

Or push from local: git remote add origin https://huggingface.co/spaces// git push --set-upstream origin main

As soon as the repository contains app.py (or main.py) and requirements.txt, the Space will auto-build and launch.