Better Call Saul - AI Corporate Law Squad

Multi-agent AI advisory system that coordinates Tax, Legal, and Strategy specialists to provide unified business formation recommendations.

---

The Problem

Starting a business? You ask three professionals what entity to form:

• Tax CPA: "Use an LLC for tax savings"
• Attorney: "Use a C-Corp for liability protection"
• Strategist: "Use a C-Corp for investor appeal"

Three experts. Three different answers. Zero clarity.

Why? Traditional consultants work in silos—each optimizes for their domain without understanding how their advice impacts the others.

---

The Solution

Better Call Saul coordinates three AI specialist agents through an intelligent orchestrator:

Tax CPA Agent - Analyzes tax implications, deductions, and strategies
Corporate Attorney Agent - Evaluates legal structures, liability, and compliance
Business Strategist Agent - Assesses growth plans, funding, and scalability

The Coordinator Agent identifies conflicts and synthesizes unified recommendations with explained trade-offs.

Result: Integrated strategy instead of conflicting opinions.

---

Architecture

User Query
    ↓
Coordinator Agent
    ├─→ Tax CPA Agent (parallel)
    ├─→ Attorney Agent (parallel)
    └─→ Strategist Agent (parallel)
         ↓
    Conflict Resolution
         ↓
    Unified Strategy

Tech Stack

Component	Technology
Frontend	React + Vite, react-markdown
Backend	Python 3.12, Google ADK
AI Model	Gemini 2.5 Flash
Deployment	Cloud Run (2 services)
Infrastructure	Cloud Build, Artifact Registry

---

Local Development

Prerequisites

• Python 3.12+
• Node.js 20+
• Google Cloud CLI
• Gemini API key (get one here)

Backend Setup

cd backend
python -m venv venv
source venv/bin/activate 

pip install google-adk python-dotenv

# Create .env file
echo "GOOGLE_API_KEY=your_key_here" > .env
echo "GOOGLE_GENAI_USE_VERTEXAI=FALSE" >> .env

# Start ADK API server
adk api_server --allow_origins="*"

Backend runs on http://localhost:8000

Frontend Setup

cd frontend
npm install
npm run dev

Frontend runs on http://localhost:5173

---

Deploy to Cloud Run

Backend Deployment

cd backend

# Deploy with auto-scaling
gcloud run deploy better-call-saul-backend \
  --source . \
  --region=us-central1 \
  --memory=2Gi \
  --cpu=2 \
  --min-instances=0 \
  --max-instances=10 \
  --allow-unauthenticated \
  --set-env-vars=GOOGLE_API_KEY=YOUR_KEY,GOOGLE_GENAI_USE_VERTEXAI=FALSE

Frontend Deployment

cd frontend

# Update backend URL in src/App.jsx
# Replace localhost:8000 with your backend Cloud Run URL

npm run build

# Deploy
gcloud run deploy better-call-saul-frontend \
  --source . \
  --region=us-central1 \
  --memory=512Mi \
  --cpu=1 \
  --min-instances=0 \
  --max-instances=10 \
  --allow-unauthenticated

---

Project Structure

better-call-saul/
├── backend/
│   ├── corporate_law_squad/
│   │   ├── agent.py           # Multi-agent definitions
│   │   └── __init__.py
│   ├── Dockerfile             # Backend container config
│   └── .env                   # API keys (gitignored)
├── frontend/
│   ├── src/
│   │   ├── App.jsx            # Main React component
│   │   ├── App.css            # Styling
│   │   └── main.jsx
│   ├── Dockerfile             # Frontend container config
│   └── nginx.conf             # Production server config
└── README.md

---

Key Features

✅ Parallel Agent Execution - 2-3x faster than sequential consultation
✅ Conflict Resolution - Identifies contradictions and synthesizes unified advice
✅ Session Management - Maintains conversation context across queries
✅ Markdown Rendering - Professional formatted output with structure
✅ Serverless Auto-scaling - Scales 0-10 instances based on traffic
✅ Cost Optimized - $0 when idle, ~$5/month for moderate usage

---

Architecture Diagrams

---

License

MIT License - see LICENSE for details