Reachout Agent - AI-Native Outreach Copilot

An AI-powered outreach assistant for freelancers and solo founders that automates business discovery, research, and personalized message generation.

Project Status

Phase: Foundation (Ralph autonomous development) Backend: Not yet implemented Docs: Complete and ZeroDB-aligned

Documentation

• PRD - Product requirements and architecture
• Data Model - ZeroDB table and embedding schemas
• Backlog - Epic-based user stories
• Sprint Plan - 1-day MVP build plan

Development Approach

This project uses Ralph autonomous development:

• Ralph autonomously implements backend following the PRD
• Each iteration is tracked in progress.txt
• Story completion tracked in prd.json

Ralph Files

• prd.json - User stories in Ralph format (generated from docs/)
• progress.txt - Iteration learnings and context
• AGENTS.md - Project-specific agent instructions
• ralph.sh - Autonomous execution loop (in ralph/ subdirectory)

Architecture

Backend: FastAPI + AI-Native Studio APIs Data Layer: ZeroDB (tables + embeddings) Agents: Discovery, Enrichment, Qualification, Drafting, Learning External Tools: Web search MCPs, Firecrawl scraper

See PRD for full architecture.

Getting Started

Once Ralph completes foundation:

cd backend
pip install -r requirements.txt
python -m uvicorn main:app --reload

Frontend lives in separate repo.