Skip to content

README.md

Latest commit

 

History

History
129 lines (88 loc) · 3.09 KB

README.md

File metadata and controls

129 lines (88 loc) · 3.09 KB
Raw

PingGenius Backend

Transform LinkedIn profile information into personalized outreach sequences.

Built with FastAPI for simple, reliable use.

The Story

In October 2025 I started building PingGenius as my first serious AI project.

I wanted to solve a real problem: most LinkedIn outreach tools either scrape data or generate generic messages that get ignored.

After months of work I learned what it takes to make an AI backend reliable:

  • avoid incorrect output
  • keep responses consistent
  • control costs and timeouts
  • make messages sound natural

I paused product work for a while to focus on my mental health, but I kept the backend alive because the engineering lessons were valuable.

This is a clean FastAPI backend. It can power a SaaS or be used as a foundation for your own outreach tool.

Core Capabilities

  • Smart profile analysis – extracts role, company, industry, pain points, and recent activity
  • Personalized sequences – connection notes, DMs, and follow-ups
  • Tone control – Friendly, Direct, Authority, Casual
  • Message refinement – improve messages while keeping the sequence consistent
  • Human-in-the-loop – you copy/paste; no auto-sending
  • Pain anchoring – find the main pain point before generating

Technical Features

  • FastAPI (async)
  • OpenAI Python SDK
  • MongoDB
  • Authentication and rate limiting
  • Swagger/OpenAPI docs
  • GDPR-compliant handling
  • Timeouts, output sanitization, and idempotency

Quick Start

Prerequisites

  • Python 3.11+
  • uv package manager (recommended)

Installation

git clone https://github.com/hasnainXdev/pinggenius_backend
cd pinggenius_backend

python -m venv venv
source venv/bin/activate    # Windows: venv\Scripts\activate

uv add -r requirements.txt

Run locally

uvicorn main:app --reload

Open http://localhost:8000/docs for Swagger UI.

API Documentation

All v1 endpoints are documented.

Key endpoints:

  • POST /api/v1/profile/analyze – analyze LinkedIn profile data
  • POST /api/v1/outreach/generate – generate outreach sequence
  • POST /api/v1/outreach/refine – refine messages
  • GET /api/v1/outreach/{id} – retrieve saved sequence

Full docs → /docs

Security & Compliance

  • no account risk (copy-paste only)
  • GDPR-compliant data handling
  • rate limiting and request validation
  • timeouts and output checks
  • idempotency to avoid duplicate processing

What I Learned

This project taught me about building an AI backend:

  • making outputs predictable
  • why many AI tools fail
  • why safety layers matter

The main issues were fixed before I considered this ready.

Development Roadmap

Must-fix (Completed)

  • guard against empty/weak profiles
  • deterministic output sanitization
  • timeout and runaway protection
  • idempotency
  • pain anchoring
  • sequence cohesion memory
  • tone drift protection

Nice-to-have

  • reply-probability scoring
  • A/B variants
  • LinkedIn policy-safe checker

Built By

Muhammad Hasnain

AI Engineer & Full-Stack Developer from Karachi, Pakistan hasnainXdev on GitHub & X

Building useful AI tools for peoples.

For developers: fork it and improve it.