Skip to content

README.md

Latest commit

 

History

History
66 lines (46 loc) · 2.96 KB

README.md

File metadata and controls

66 lines (46 loc) · 2.96 KB

Technical Documentation

This directory contains deep implementation details for the AI GTM Stakeholder Engine's hybrid cloud architecture combining local AI processing, cloud vector database, and global frontend access.

🏗️ Technical Overview

System Type: Hybrid cloud architecture with local-to-global data flow Core Stack: Python (11-agent pipeline), Vector database + RAG, React/TypeScript (global interface) Architecture Pattern: Local AI enhancement → Cloud knowledge storage → Worldwide team access Integration: Anthropic Claude API, OpenAI embeddings, pgvector-compatible database, Docker deployment

🎯 Developer Navigation by Technical Need

"I need complete system understanding"

Path: agent_specifications.md → data_store_specification.md → diagrams/ Purpose: Architecture review, system evaluation, technical due diligence

"I'm implementing/deploying the system"

Path: deployment_guide.md → agent_specifications.md → data_store_specification.md Purpose: Production deployment, system setup, operational configuration

"I'm extending/contributing to the system"

Path: agent_specifications.md → data_store_specification.md → ../contrib/ Purpose: Feature development, agent creation, system enhancement

"I need visual system overview"

Path: diagrams/ → agent_specifications.md Purpose: Presentations, system demos, architectural discussions

📋 Documentation Inventory

System Architecture & Design

  • agent_specifications.md → 11-agent pipeline execution and RAG system integration
  • data_store_specification.md → Multi-project isolation and vector database architecture

Implementation & Deployment

  • deployment_guide.md → Production deployment procedures and system setup

Visual Reference

  • diagrams/ → System flowcharts, pipeline execution flow, file dependencies

🔧 Architecture Summary

3-Layer Hybrid System:

  • Local Processing: 11 AI agents enhance strategic assets privately (Python + Claude API)
  • Cloud Storage: Vector database with semantic search (pgvector + OpenAI embeddings)
  • Global Access: Team interface with department views (React + TypeScript)

Key Integration Points:

  • Agent 11 (RAG System) bridges local→cloud data flow
  • Frontend queries cloud database for instant strategic guidance
  • Multi-project isolation with shared learning system

⚙️ Implementation Prerequisites

Development Environment: Python 3.8+, Node.js 18+, Docker (optional) API Requirements: Anthropic API key, OpenAI API key, vector database with pgvector support System Resources: 4GB RAM, 1GB storage per project Network: API access with retry capability

Quick Setup Reference: See deployment_guide.md for full-stack deployment Complete Deployment: See deployment_guide.md for backend pipeline + frontend setup

For user-facing setup and business documentation, see ../docs/