Skip to content

rishi-banerjee1/Sourcing_Compass

Repository files navigation

Sourcing Compass

AI-powered talent landscape mapper — describe a role and get target companies, adjacent pools, wildcard bets, and job titles in seconds.

React Groq GitHub Pages License


Why This Exists

  • Recruiters spend hours manually researching where talent comes from before sourcing begins
  • Standard ATS/LinkedIn workflows give you job titles, not talent ecosystems
  • "Adjacent pools" and "wildcard bets" (e.g. consulting firms, research labs) are invisible without domain knowledge
  • One form → AI-generated map replaces a 2-hour desk research session
  • Connections between nodes (companies ↔ titles ↔ pools) are visualised on hover — not buried in a spreadsheet

How to Use It

Live App

Go to https://rishi-banerjee1.github.io/Sourcing_Compass/

No login. No account. Bring your own Groq API key (free at console.groq.com).


Step-by-step

1 — Fill in the left panel

Field Required Example What it does
Role Title ✅ Yes Staff Backend Engineer Core search anchor — the AI reasons from this
Hiring Company Optional Stripe Narrows company peer set to relevant industry
Location Optional Singapore Filters for geo-relevant talent pools
Seniority Optional Senior / Staff Calibrates scope and depth of results
Key Skills Optional Rust, distributed systems Adds skill-specific company/title suggestions
Industries Optional Fintech, Infrastructure Keeps adjacent pools on-target
Exclusions Optional Google, Meta Suppresses companies you don't want to source from

All comma-separated fields accept free text — type and hit Enter or comma to add a tag.

2 — Click Generate Map

The AI (Llama 3.3 70B via Groq) processes your inputs and returns a structured talent landscape. Takes 5–10 seconds.

3 — Read the four sections

Section What's in it How to use it
Target Companies Companies most likely to employ this profile Start sourcing here first
Adjacent Talent Pools Consulting firms, research labs, adjacent verticals Use for passive candidates who wouldn't show on direct title search
Wildcard Bets Non-obvious, high-signal sources Useful for niche roles or when the obvious pools are exhausted
Target Titles Exact and variant job titles with match confidence % Paste these into LinkedIn/Ashby searches

Each card shows:

  • Company/title name + sub-label
  • Hiring stage badge (if applicable)
  • Tags (tech stack, industry signals)
  • Number of cross-connections to other nodes

4 — Hover to reveal connections

Hover over any card to see dashed blue lines connecting it to related nodes across all four sections. Use this to understand which companies map to which titles, or which adjacent pool feeds which company cluster.

5 — Iterate

Refine your inputs — add exclusions, change seniority, narrow industry — and click Generate Map again. Each run is independent.


Self-Hosting / Local Dev

You need a free Groq API key.

git clone https://github.com/rishi-banerjee1/Sourcing_Compass.git
cd Sourcing_Compass
npm install
cp .env.example .env          # add your VITE_GROQ_API_KEY
npm run dev                   # starts on http://localhost:5173

The API key is injected at build time via Vite's import.meta.env. It is embedded in the browser bundle — do not use a key with billing limits you aren't comfortable exposing client-side. Groq's free tier is suitable for personal use.

Deploy your own GitHub Pages fork

  1. Fork the repo
  2. Go to Settings → Secrets → Actions and add VITE_GROQ_API_KEY
  3. Go to Settings → Pages, set source to GitHub Actions
  4. Push to main — the workflow deploys automatically

Tech Stack

Layer Choice Why
UI React 18 + Tailwind CSS v4 Component model, fast iteration
Build Vite 5 Zero-config, fast HMR
AI Groq (Llama 3.3 70B) Fast inference, generous free tier
Hosting GitHub Pages via Actions Zero cost, no server needed

Contributing

PRs welcome — see SECURITY.md for reporting vulnerabilities.

For feature changes, open an issue first to align on scope.

License

MIT

About

AI-powered talent landscape mapper for recruiters — generates target companies, adjacent pools, wildcard bets, and job titles from a single role input. Built with React + Groq (Llama 3.3 70B).

Topics

Resources

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages