This project analyzes job descriptions to extract and track trending technical skills over time using automation and AI.

• Automate ingestion of job description files from Google Drive
• Extract relevant technical skills from each job description
• Index the extracted data using ChromaDB (vector store)
• Maintain a Google Sheet for skill frequency and trends
• Visualize top skills per job type or time period

• Flask: Simple local UI for job description submission and status tracking
• n8n: No-code/low-code automation platform (Docker)
• Python: For skill extraction, processing, and Chroma integration
• ChromaDB: Lightweight vector database for skill indexing
• Google Drive & Sheets APIs: Input and output data
• OpenAI API: Skill extraction using LLMs

This project is under active development.
Final directory structure and setup instructions will be documented here once the architecture is stable.

CURRENTLY EVOLVING!!!

• Simple input form with two tabs
  • Tab 1 – Submission
    • Input fields for:
      • Job Title
      • Company
      • Job Description
    • Header shows the count of pending and processed files
    • On submission, a new text file named with a title/timestamp is saved to a Pending folder in Google Drive
  • Tab 2 – Status
    • Show a table of uploaded job files with:
      • File name
      • Timestamp
      • Status – pending or processed
    • Files do not need to be viewable in the first iteration

• Poll Google Drive Pending folder on a schedule
• For each file:
  1. Get job title and description
  2. Use OpenAI API to determine skills from the description text
  3. Save skills as structured data
  4. Index these skills in ChromaDB (vector store)
  5. Move the file to the Google Drive Processed folder

• Enable retrieval of the top 15 skills for a given job title using similarity search or tag aggregation (TBD)
• Process:
  1. Python script running on cron or n8n scheduled task
  2. Use ChromaDB API to run similarity search on stored data
  3. Filter by job titles or embedding similarity
  4. Aggregate skills
  5. Return the top 15 with counts:
     • Number of occurrences
     • Number of jobs (weighted)
• Output formats: JSON, CSV, or similar for stretch goals

1. Visualization: Display top skills (bar chart, heatmap, etc.)
2. Trends Over Time: Compare skill demands for a given title now vs. 6 months later to track evolution

• No job scraping – respect TOS of job boards
• No authentication in the initial version beyond required API keys (single-user app, running locally)
• Visualization is optional – raw CSV/text output is fine for now
• Learning resources included ChatGPT, YouTube, and Udemy.

• Google Sheet skill trend heatmap
• Dashboards or Streamlit visualizations
• Webhook-based ingestion pipeline

Let me know if you'd like to customize this with your blog, LinkedIn, or repo badges.

Once you're happy with this, you can:

git add README.md
git commit -m "Add initial README with project goals and setup"
git push