Skip to content

SFAI SDK Documentation

Welcome to the SFAI SDK! The complete toolkit for taking your AI applications from local development all the way to Salesforce AgentForce integration. Build sophisticated AI services with multiple framework support, test changes instantly with Docker-based local deployment, deploy seamlessly to production-ready platforms, and connect directly to Salesforce AgentForce through MuleSoft with one unified toolkit.

🎯 One SDK, Complete Journey

From sfai app init to AgentForce — Experience the fastest path from idea to production-ready AI agents in the Salesforce ecosystem.

✨ Zero complexity. Maximum impact.

🚧 ACTIVE DEVELOPMENT NOTICE

SFAI SDK is currently in active development. Features and APIs may change as we continue to enhance the platform.

If you encounter any issues, please open a GitHub issue or submit a pull request. Feedback and contributions are very welcome!

🚀 Key Features

🐳 Local Development

Zero-config Docker deployment for instant testing

sfai app init     # 📦 Scaffold your app
sfai app deploy   # 🚀 Deploy locally
sfai app open     # 🌐 Open in browser

✨ What you get:

  • Zero-config local deployment: Simple Docker containerization for instant testing
  • Automatic port management: Finds available ports and handles networking
  • Hot reloading: Test changes immediately without complex setup
  • Consistent environments: Same container runs locally and in production-ready platforms

Enterprise-grade platforms ready for scale

sfai platform init --cloud heroku  # ⚙️ Configure platform
sfai app deploy                     # 🚀 Deploy to production

🌟 Supported Platforms:

  • Heroku → Native Salesforce platform integration with Private Spaces support
  • Amazon EKS → Kubernetes deployment with auto-scaling and enterprise features
  • More platforms coming soon → Expanding cloud provider support

Multiple frameworks, same seamless experience

sfai app init --template fastapi_hello  # 🐍 FastAPI (Available Now)
sfai app init --template flask_hello     # 🌶️ Flask (Coming Soon)

🔧 Current & Upcoming:

  • FastAPI ✅ Currently supported with full Docker + Helm integration
  • Flask 🔜 Coming soon with the same seamless deployment experience
  • Additional frameworks 🚀 Expanding template library for diverse use cases

Production-ready from day one

sfai app publish --service mulesoft  # 🔗 Publish to MuleSoft
sfai app context                     # 📊 View all configurations

💼 Enterprise-Ready:

  • Persistent context system → Manages all app configuration and deployment state
  • MuleSoft API integration → Publish and deploy APIs for Salesforce connectivity
  • Simplified environment management → Switch between local, staging, and production
  • Zero configuration → Context system handles all environment settings automatically

🏆 Why Choose SFAI SDK

🚀 From Local to AgentForce with one SDK

SFAI SDK provides the only toolkit you need for the complete AI application lifecycle in the Salesforce ecosystem

The fastest path from idea to production-ready AgentForce integration

Step-by-Step Journey

1️⃣ Start Locally 

sfai app init && sfai app deploy
# 🎉 Your app is running at http://localhost:8080

2️⃣ Deploy to Production-Ready Platform 

sfai platform init --cloud heroku && sfai app deploy
# 🎉 Your app is live on Heroku with Salesforce integration

3️⃣ Connect to Salesforce 

sfai app publish --service mulesoft
# 🎉 Your app is now an AgentForce action available to any agent!

✨ Key Benefits:

  • One command deployment → Same codebase runs everywhere from your laptop to Salesforce AgentForce
  • Unified context management → All configurations, credentials, and deployments managed in one place
  • Zero friction scaling → Seamless transition from prototype to enterprise production-ready platforms

Why developers choose SFAI SDK over traditional approaches

With SFAI SDK

Simple. Fast. Integrated. 

sfai app init           # One tool
sfai app deploy         # Same commands
sfai app publish        # Consistent workflow
Result: 🎉 Your app is available as an action in AgentForce

Traditional Approach

Complex. Slow. Fragmented. 

# Multiple tools, complex setup
docker build && kubectl apply && helm install
# + MuleSoft setup + Salesforce configuration
# + Environment management + Credential juggling
Result: 😵 Weeks of integration work!

🌟 Why Developers Love It:

  • 🛠️ Single SDK → No need to learn multiple tools, platforms, or deployment processes
  • 🔄 Consistent Workflow → Same commands work for local testing and production platform deployment
  • ⚡ Instant Feedback → Test your ai apps locally before deploying to production platforms
  • 🧠 Context Aware → Remembers your preferences and configurations across all environments

Salesforce Ecosystem Integration

SFAI SDK is designed to work natively with the Salesforce ecosystem:

  • Heroku Deployment: Deploy directly to Salesforce's preferred application platform with enterprise features like Private Spaces and internal routing
  • MuleSoft Connectivity: Publish and deploy APIs to MuleSoft for seamless Salesforce integration
  • AgentForce Integration: Direct pathway from your AI service to Salesforce AgentForce capabilities
  • Enterprise Security: Support for Salesforce SSO and secure network configurations
  • Scalable Architecture: Built for enterprise-grade applications with high availability

Ready to build? Follow the Installation Guide to get started, or jump to the Quickstart for immediate deployment!

Need production deployment? Check out our Heroku Deployment and Local Development guides.