Skip to main content
Deploy your agents as APIs, web apps, or serverless functions to share with others.

​
Deployment Options

Local Script

Run directly on your machine

Web API

Flask, FastAPI, or similar

Serverless

AWS Lambda, Cloud Functions

Built-in UI

PraisonAI’s Chainlit interface

​
Quick Start: Built-in UI

The easiest way to deploy with a chat interface: 
pip install praisonai
praisonai ui
This launches a Chainlit-based chat interface for your agents.

​
Web API with FastAPI

from fastapi import FastAPI
from praisonaiagents import Agent

app = FastAPI()

# Create agent once at startup
agent = Agent(
    name="SupportBot",
    instructions="You are a helpful support agent"
)

@app.post("/chat")
async def chat(query: str):
    response = agent.start(query)
    return {"response": response}

# Run with: uvicorn app:app --reload

​
Web API with Flask

from flask import Flask, request, jsonify
from praisonaiagents import Agent

app = Flask(__name__)

agent = Agent(
    name="SupportBot",
    instructions="You are a helpful support agent"
)

@app.route("/chat", methods=["POST"])
def chat():
    query = request.json.get("query", "")
    response = agent.start(query)
    return jsonify({"response": response})

# Run with: python app.py
if __name__ == "__main__":
    app.run(port=5000)

​
Serverless (AWS Lambda)

import json
from praisonaiagents import Agent

agent = Agent(
    name="ServerlessAgent",
    instructions="You are a helpful assistant"
)

def lambda_handler(event, context):
    body = json.loads(event.get("body", "{}"))
    query = body.get("query", "")
    
    response = agent.start(query)
    
    return {
        "statusCode": 200,
        "body": json.dumps({"response": response})
    }

​
Environment Variables

Always use environment variables for API keys: 
export OPENAI_API_KEY=your_key_here

import os
from praisonaiagents import Agent

# API key is automatically read from environment
agent = Agent(instructions="You are helpful")
Never hardcode API keys in your code. Use environment variables or secrets managers.

​
Scaling Considerations

Rate Limiting

Control API usage and costs

Caching

Cache common responses

Error Handling

Handle API failures gracefully

Monitoring

Track usage and performance

​
Best Practices

Verify agent works correctly with various inputs
Never hardcode API keys or secrets
Wrap agent calls in try/except blocks
Track API calls and costs

​
Course Complete! 🎉

You’ve learned how to:
  • ✅ Create AI agents with instructions
  • ✅ Add tools for extended capabilities
  • ✅ Enable memory for conversations
  • ✅ Add knowledge from documents
  • ✅ Build multi-agent teams
  • ✅ Create specialized agents
  • ✅ Deploy agents for real use

​
Next Steps

Explore Docs

Deep dive into advanced features

Join Community

Connect with other builders

View Examples

See real-world implementations

Build Something

Apply what you’ve learned!