Skip to content

MOREQASSE/LEADFinder

Repository files navigation


LEADFinder

LEADFinder
By Devaxio

AI-powered lead discovery engine — scrapes, ranks, and drafts replies across
8+ platforms so you never miss a client.


Python FastAPI React Tailwind CSS SQLite OpenRouter AI



LEADFinder is a full-stack localhost application that turns the chaos of 8 different platforms into one clean, brutalist dashboard. It scrapes Reddit, Craigslist, Upwork, Google Alerts, Mastodon, Indeed, and Hacker News — then uses AI to rank leads by budget/urgency and draft personalized replies from your uploaded resume.

Zero monthly cost. Runs on your machine. Scales with your team.




LEADFinder Live Pipeline
/// live pipeline — jobs streaming in from 8 platforms, ranked by AI ///




🚀 Quick Start

Python (Windows — first time)

# 1. Download Python from https://www.python.org/downloads/ (check "Add Python to PATH")
# 2. Verify it's installed:
python --version
# 3. Upgrade pip (optional but recommended):
python -m pip install --upgrade pip

Python (Linux — first time)

# Debian / Ubuntu / Mint:
sudo apt update && sudo apt install python3 python3-pip python3-venv -y

# Fedora:
sudo dnf install python3 python3-pip -y

# Arch:
sudo pacman -S python python-pip --noconfirm

# Verify:
python3 --version && pip3 --version

Project Setup

# 1. Install backend dependencies
pip install -r requirements.txt

# 2. Install frontend dependencies
cd frontend && npm install && cd ..

# 3. Launch both servers
python run.py

Then open http://localhost:5173 — the backend API docs live at http://localhost:8000/docs.

Tip

First run? Create an account, upload your resume (AI auto-parses it), then head to Settings → Presets to pick a search configuration. Full workflow in the Launch Guide.




🔧 Tech Stack

Python JavaScript HTML5 CSS3

Layer Technologies
Backend
AI / LLM
Frontend
Scraping
Infra



📂 Structure

LEADFinder/
├── backend/                    # FastAPI application
│   ├── ai/                     # AI clients + draft generator
│   │   ├── openrouter_client.py
│   │   ├── github_client.py
│   │   └── draft_generator.py
│   ├── routers/                # API endpoints
│   │   ├── auth.py             # Login / register / OAuth
│   │   ├── leads.py            # CRUD + ranking
│   │   ├── replies.py          # Drafts + send + retry
│   │   ├── settings.py         # Budgets / limits / keywords
│   │   └── ai_config.py        # Model provider config
│   ├── scrapers/               # Source-specific scrapers
│   │   ├── reddit.py
│   │   ├── craigslist.py
│   │   ├── upwork.py
│   │   ├── mastodon.py
│   │   ├── indeed.py
│   │   ├── hackernews.py
│   │   └── google_alerts.py
│   ├── utils/                  # Ranking, currency formatting
│   ├── main.py                 # FastAPI entry point
│   ├── database.py             # SQLAlchemy config
│   └── models.py               # DB schemas
│
├── frontend/                   # React application
│   ├── src/
│   │   ├── components/         # UI components
│   │   │   ├── Auth/
│   │   │   ├── Dashboard/
│   │   │   ├── Layout/         # Sidebar, Header
│   │   │   ├── Leads/
│   │   │   ├── Settings/       # Tabbed settings page
│   │   │   └── Profile/
│   │   ├── hooks/              # useAuth, useSettings, useLeads
│   │   └── services/
│   │       └── api.js          # Axios instance
│   ├── package.json
│   └── vite.config.js
│
├── assets/                     # Images, screenshots
├── requirements.txt            # Python dependencies
└── run.py                      # One-command launcher



🎯 Features

Multi-Platform Scraping
8 sources — Reddit, Craigslist, Upwork, Google Alerts, Mastodon, Indeed, Hacker News. Each runs on its own schedule with configurable rate limits.
AI-Powered Ranking
Leads are auto-ranked Hot / Warm / Cold based on budget keywords and urgency signals. Budgets converted to MAD for comparison.
Resume-Aware Drafting
Upload your resume once — AI parses your name, portfolio, pricing, skills, and tone. Every draft is personalized and on-brand.
Kanban & List Views
Toggle between chronological list, Kanban board (New → Drafted → Replied → Archived), or split view for high-volume triage.
Configurable Everything
Budget ranges, rate limits, keyword boosts/blacklists, AI models, search presets — all editable from the Settings panel.
Multi-User Ready
1 admin + 2-3 team members. OAuth (Google/GitHub) or email/password auth. Local SQLite — no cloud costs.



🏁 First Launch

  1. Open localhost:5173 and create an admin account
  2. Go to Profile and upload your resume — AI parses it instantly
  3. Go to Settings → Presets and pick a search preset (or create your own)
  4. Configure Connection Center with your API keys (Reddit, Gmail, etc.)
  5. Set Budget & Limits to match your workflow
  6. Dashboard populates automatically as scrapers find leads



📷 Screenshots

⚡ Full Dashboard 🔨 17+ Built-in Tools
Dashboard Tools Page
Rank, filter, bulk-edit, and track every lead from one brutalist cockpit Word counter, QR generator, JSON formatter, invoice generator & more — all free
📄 ATS Resume Checker 🧾 Invoice Generator
ATS Resume Checker Invoice Generator
Paste any job description — Our Algorithm scores your resume and highlights skill gaps Create branded PDF invoices with one click, ready to send



Built with by Devaxio
LEADFinder — 2026 — Local-first — AI-native

About

LEADFinder is a full-stack localhost application that turns the chaos of 8 different platforms into one clean, brutalist dashboard. It scrapes Reddit, Craigslist, Upwork, Google Alerts, Mastodon, Indeed, and Hacker News — then uses AI to rank leads by budget/urgency and draft personalized replies from your uploaded resume. Zero monthly cost.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors