Deployment Overview

Options for deploying agents to production

Deploy your PraisonAI agents to production environments.

from praisonaiagents import Agent

agent = Agent(
    name="Production Agent",
    instructions="Handle production workloads.",
)

agent.start("Ready for production deployment")

The user chooses a hosting pattern, configures secrets and scaling, then points clients at the running agent.

Quick Start

1. Simple Usage

   Start with an agent-centric approach:

   from praisonaiagents import Agent
   
   # Create production-ready agent
   agent = Agent(
       name="Production Agent",
       instructions="Handle production workloads",
   )
   
   # Deploy to production
   agent.start("Ready for production deployment")
   
2. With Configuration

   Choose between different deployment options based on your needs.

Deployment Methods

24/7 Production Setup

Start here for always-on deployment with process supervision

Advanced Production

Comprehensive production guide with monitoring and scaling

Deployment Options

Option	Best For	Complexity	Guide
Production Minimal	Always-on 24/7 operation	Low	📖 Guide
Docker Compose	Containerized deployments	Low	📖 Guide
Kubernetes	Scalable production	Medium	📖 Guide
Cloud Functions	Serverless	Low	📖 Guide
API Server	REST/WebSocket APIs	Medium	📖 Guide

Quick Deploy with CLI

# Deploy as API server
praisonai deploy --type api

# Deploy as Docker container
praisonai deploy --type docker

# Deploy to cloud
praisonai deploy --type cloud --provider aws

API Server

from praisonaiagents import Agent
from praisonaiagents.api import serve

agent = Agent(name="API Agent")

# Start API server
serve(agent, host="0.0.0.0", port=8000)

Environment Variables

# Required
export OPENAI_API_KEY="sk-..."

# Optional
export PRAISONAI_LOG_LEVEL="INFO"
export PRAISONAI_DB_URL="postgresql://..."

Related

Production Minimal

Always-on deployment

Docker Deployment

Containerised agents