Deploying Agents

How to deploy and share your AI agents with others

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

Diagram

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:

bash

pip install praisonai
praisonai ui

This launches a Chainlit-based chat interface for your agents.

Web API with FastAPI

python

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

python

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)

python

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:

bash

export OPENAI_API_KEY=your_key_here

python

import os
from praisonaiagents import Agent

# API key is automatically read from environment
agent = Agent(instructions="You are helpful")

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

Test Before Deploying

Verify agent works correctly with various inputs

Use Environment Variables

Never hardcode API keys or secrets

Add Error Handling

Wrap agent calls in try/except blocks

Monitor Usage

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!