VectorOS - AI-Powered Revenue Intelligence Platform

The AI that stops deals from dying - Autonomously monitors your pipeline, learns from every outcome, and prevents revenue loss before it happens.

✅ Current Status: AUTONOMOUS INTELLIGENCE LIVE!

Core Platform: ✅ Operational AI Insights: ✅ Claude Sonnet 4.5 Integration Autonomous Monitoring: ✅ NEW! Always-On Intelligence Engine

What's New:

🤖 Continuous Monitoring Worker - Analyzes all deals every 30 minutes automatically
🎯 Intelligent Anomaly Detection - Detects stale deals, velocity drops, risk signals
🚨 Proactive Insights - AI generates alerts without manual intervention
📊 Deal-First Insights Dashboard - Scalable for 5000+ deals

This is true Revenue Intelligence, not just a CRM with AI.

See: AUTONOMOUS_MONITORING_LIVE.md for details.

Quick Start

Prerequisites

Node.js >= 18.0.0
Python >= 3.10
Neon PostgreSQL account (or compatible PostgreSQL)
Anthropic API key (Claude)

Installation

# Clone the repository
git clone https://github.com/ptengelmann/VectorOS.git
cd VectorOS
# Install frontend dependenciescd frontend && npm install &&cd ..
# Install backend dependenciescd backend && npm install &&cd ..
# Install AI Core dependenciescd ai-core
python3 -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -r requirements.txt
cd ..
# Set up environment variables# Create .env in project root with:
DATABASE_URL="postgresql://..."
ANTHROPIC_API_KEY="sk-ant-..."
NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY="pk_test_..."
CLERK_SECRET_KEY="sk_test_..."# Create symlinks for servicescd backend && ln -sf ../.env .env &&cd ..
cd ai-core && ln -sf ../.env .env &&cd ..
# Initialize databasecd backend
npx prisma generate
npx prisma db push
cd ..

Run Development Servers

Terminal 1 - Frontend:

cd frontend
npm run dev
# Runs on http://localhost:3000

Terminal 2 - Backend:

cd backend
npm run dev
# Runs on http://localhost:3001

Terminal 3 - AI Core:

cd ai-core
source venv/bin/activate
python -m src.main
# Runs on http://localhost:8000

Verify Installation

Frontend: http://localhost:3000
Backend Health: http://localhost:3001/health
AI Core Health: http://localhost:8000/health

Quick Start with Autonomous Monitoring

# One-command startup (all services + autonomous worker)
./start-autonomous.sh
# Stop all services
./stop-autonomous.sh

Project Structure

VectorOS/
├── frontend/ # Next.js 16 + TypeScript + Tailwind CSS
│ ├── app/ # App router pages
│ ├── components/ # React components
│ └── lib/ # API client and utilities
│
├── backend/ # Node.js + Express + Prisma
│ ├── src/
│ │ ├── services/ # Business logic services
│ │ └── index.ts # API routes
│ └── prisma/ # Database schema
│
├── ai-core/ # Python + FastAPI + Claude Sonnet 4
│ ├── src/
│ │ ├── services/ # AI services (insights, scoring, analysis)
│ │ ├── config.py # Configuration
│ │ └── main.py # FastAPI app
│ └── requirements.txt
│
├── docs/ # Architecture and planning docs
└── .env # Environment variables (create this)

Tech Stack

Frontend

Framework: Next.js 16 with App Router
Language: TypeScript
Styling: Tailwind CSS
Auth: Clerk
State: React Hooks + localStorage

Backend

Runtime: Node.js
Framework: Express.js
ORM: Prisma
Database: Neon PostgreSQL (serverless)
Language: TypeScript

AI Core

Framework: FastAPI
Language: Python 3.10+
LLM: Claude Sonnet 4 (Anthropic API)
Embeddings: sentence-transformers (all-MiniLM-L6-v2)
Vector DB: Qdrant (planned)

Features

Current (Implemented)

✅ Authentication: Clerk integration, workspace management
✅ Deal Management: Create, view, update deals with health scoring
✅ Dashboard: Real-time metrics and pipeline visualization
✅ AI Insights (RAG-based): Intelligent insights generator using Claude Sonnet 4
- Retrieval-Augmented Generation with vector memory
- Context-aware deal analysis
- Structured insights with confidence scores
✅ AI Deal Analysis: Deep individual deal analysis with Claude
✅ Health Scoring: Algorithmic deal health scoring (6 dimensions)

In Progress

🔄 Insights Generation Refinement: Optimizing Claude prompts for better results
🔄 Vector Database Integration: Full Qdrant setup for semantic search

Planned (Next 4-12 Weeks)

📅 Revenue Forecasting: AI-powered revenue predictions with confidence intervals
📅 Pipeline Health Dashboard: Real-time bottleneck detection
📅 Churn Prediction: Prevent customer churn before it happens
📅 Email/Calendar Integration: Automatic deal capture from Gmail/Outlook
📅 Autonomous Monitoring: 24/7 background monitoring with automatic alerts
📅 Advanced Analytics: Rep performance, cycle time analysis

API Endpoints

Backend API (Port 3001)

Authentication & Users
POST /api/v1/users Create user
GET /api/v1/users/:userId Get user
Workspaces
POST /api/v1/workspaces Create workspace
GET /api/v1/workspaces/:id Get workspace
Deals
GET /api/v1/workspaces/:id/deals List deals
POST /api/v1/deals Create deal
GET /api/v1/deals/:id Get deal
PUT /api/v1/deals/:id Update deal
Insights
GET /api/v1/workspaces/:id/insights List insights
POST /api/v1/workspaces/:id/insights/generate Generate insights
Health
GET /health Health check

AI Core API (Port 8000)

Insights
POST /api/v1/insights/generate Generate workspace insights
Deal Analysis
POST /api/v1/deals/analyze Analyze individual deal
POST /api/v1/deals/score Score deal health
Health
GET /health Health check

Database Schema

Core Models

User - User accounts with Clerk integration
Workspace - Multi-tenant workspaces
WorkspaceMember - User-workspace relationships with roles
Deal - Sales opportunities
- Fields: title, company, value, stage, probability, contactName, contactEmail, closeDate
- Timestamps: createdAt, updatedAt
- Relations: workspace, activities, insights
Activity - Deal interactions (emails, calls, meetings)
Insight - AI-generated alerts and recommendations
- Types: risk, opportunity, warning, recommendation
- Priorities: critical, high, medium, low
- Fields: title, description, confidence, data, actions

Documentation

📚 Core Documentation (3 Files)

COMPLETE_GUIDE.md - 📖 Everything: deployment, architecture, status, roadmap, troubleshooting
AUTONOMOUS_MONITORING.md - 🤖 Deep dive on autonomous monitoring system
README.md - ⚡ This file: quick start and overview

📝 Additional Resources

AI_INSIGHTS_ARCHITECTURE.md - 🧠 RAG architecture details
ROADMAP_AND_VISION.md - 🎯 Product strategy
POSITIONING.md - 💡 Competitive positioning

Environment Variables

Create a .env file in the project root:

# Database (Neon PostgreSQL)
DATABASE_URL="postgresql://username:password@host/database"# AI (Anthropic Claude)
ANTHROPIC_API_KEY="sk-ant-api03-..."# Auth (Clerk)
NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY="pk_test_..."
CLERK_SECRET_KEY="sk_test_..."
NEXT_PUBLIC_CLERK_SIGN_IN_URL="/sign-in"
NEXT_PUBLIC_CLERK_SIGN_UP_URL="/sign-up"
NEXT_PUBLIC_CLERK_AFTER_SIGN_IN_URL="/onboarding"
NEXT_PUBLIC_CLERK_AFTER_SIGN_UP_URL="/onboarding"# API URLs
NEXT_PUBLIC_API_URL="http://localhost:3001"
AI_CORE_URL="http://localhost:8000"

Development Workflow

Common Commands

# Kill processes on ports (if needed)
lsof -ti:3000 | xargs kill -9 # Frontend
lsof -ti:3001 | xargs kill -9 # Backend
lsof -ti:8000 | xargs kill -9 # AI Core# Database operationscd backend
npx prisma studio # Visual database browser
npx prisma generate # Regenerate Prisma client after schema changes
npx prisma db push # Push schema changes to database# View logs# Frontend: Browser console# Backend: Terminal output# AI Core: Terminal output

Making Changes

Frontend Changes:
- Edit files in frontend/app/ or frontend/components/
- Hot reload is automatic
- Check browser console for errors
Backend Changes:
- Edit files in backend/src/
- Server auto-restarts with nodemon
- Check terminal for errors
AI Core Changes:
- Edit files in ai-core/src/
- Restart the service manually
- Check terminal for errors

Database Schema Changes:

cd backend
# Edit prisma/schema.prisma
npx prisma db push
npx prisma generate

Testing

Manual Testing

Sign up / Sign in (Clerk authentication)
Create workspace (onboarding flow)
Create deals
View dashboard
Generate insights (AI Core integration)

API Testing

Use curl or Postman:

# Health checks
curl http://localhost:3001/health
curl http://localhost:8000/health
# Create deal
curl -X POST http://localhost:3001/api/v1/deals \
-H "Content-Type: application/json" \
-d '{ "title": "Acme Corp - Enterprise", "value": 120000, "stage": "negotiation", "workspaceId": "..." }'# Generate insights
curl -X POST http://localhost:3001/api/v1/workspaces/{workspaceId}/insights/generate

Troubleshooting

Backend won't start - Database connection error

Check DATABASE_URL in .env
Verify Neon database is active
Ensure .env symlink exists in backend/

AI Core authentication error

Check ANTHROPIC_API_KEY in .env
Verify key starts with "sk-ant-api03-"
Ensure .env symlink exists in ai-core/

Port already in use

# Kill the process using the port
lsof -ti:3000 | xargs kill -9

Frontend can't connect to backend

Verify backend is running on port 3001
Check NEXT_PUBLIC_API_URL in .env
Check browser console for CORS errors

Production Deployment

VectorOS is production-ready with full deployment configurations for Railway, Render, Fly.io, and more.

Quick Deploy to Railway (Recommended)

# Install Railway CLI
npm i -g @railway/cli
railway login
# Deploy (includes autonomous worker with cron)
railway init
railway up

Deploy to Render

Connect GitHub repository
Create 4 services:
- Frontend (Next.js)
- Backend (Node.js)
- AI Core (Python)
- Worker (Cron Job - every 30 minutes)

Deploy with Docker

# Build all services
docker-compose up -d
# Check logs
docker-compose logs -f worker

Complete Deployment Guide

See DEPLOYMENT_GUIDE.md for:

Step-by-step instructions for all platforms
Environment variables setup
Autonomous worker configuration
Monitoring and alerting
Scaling considerations
Cost estimates

Architecture

Three-Tier Architecture

┌─────────────────┐
│ Frontend │ Next.js (Port 3000)
│ (React/TS) │ - User interface
│ │ - Clerk authentication
└────────┬────────┘
│ REST API
▼
┌─────────────────┐
│ Backend │ Express (Port 3001)
│ (Node/TS/DB) │ - Business logic
│ │ - Database operations
└────────┬────────┘
│ REST API
▼
┌─────────────────┐
│ AI Core │ FastAPI (Port 8000)
│ (Python/ML) │ - Claude Sonnet 4 integration
│ │ - RAG-based insights
│ │ - Vector embeddings
└─────────────────┘

Product Vision

VectorOS is an AI-powered Revenue Intelligence Platform for B2B SaaS companies.

We prevent revenue loss by:

Autonomously monitoring pipelines 24/7
Predicting problems before they happen (stale deals, churn risk)
Providing actionable recommendations with AI
Learning from every outcome to improve accuracy over time

Target Market: B2B SaaS companies ($5M-$50M ARR) with 10-50 sales reps

Competitive Edge:

True autonomous monitoring (not just dashboards)
Learning system that improves with usage
Mid-market pricing ($500-3K/month vs $50K+ enterprise tools)
AI-first architecture with Claude Sonnet 4

Contributing

This is a proprietary project. For development questions or issues, contact the team.

License

Team Resources

GitHub:https://github.com/ptengelmann/VectorOS
Documentation: See /docs folder
Status Updates: See CURRENT_STATUS.md
Architecture: See AI_INSIGHTS_ARCHITECTURE.md and AUTONOMOUS_ARCHITECTURE.md

Last Updated: November 4, 2025

Name		Name	Last commit message	Last commit date
ai-core		ai-core
backend		backend
docs		docs
frontend		frontend
.env.example		.env.example
.gitignore		.gitignore
AI_INSIGHTS_ARCHITECTURE.md		AI_INSIGHTS_ARCHITECTURE.md
AUTONOMOUS_ARCHITECTURE.md		AUTONOMOUS_ARCHITECTURE.md
AUTONOMOUS_MONITORING.md		AUTONOMOUS_MONITORING.md
COMPLETE_GUIDE.md		COMPLETE_GUIDE.md
POSITIONING.md		POSITIONING.md
PRODUCTION_CHECKLIST.md		PRODUCTION_CHECKLIST.md
README.md		README.md
ROADMAP_AND_VISION.md		ROADMAP_AND_VISION.md
docker-compose.yml		docker-compose.yml
package-lock.json		package-lock.json
package.json		package.json
railway.json		railway.json
start-autonomous.sh		start-autonomous.sh
stop-autonomous.sh		stop-autonomous.sh

Latest commit

History

Folders and files

Repository files navigation