Platform-as-a-Service (PaaS) Deployment

Deploy TealTiger on managed PaaS platforms for rapid deployment without infrastructure management. This guide covers Heroku, Railway, Render, Fly.io, and other popular PaaS platforms.

Why PaaS?

Benefits:

✅ Rapid deployment - Deploy in minutes with one command
✅ Managed infrastructure - No servers, databases, or networking to manage
✅ Built-in CI/CD - Automatic deployments from Git
✅ Quick scaling - Scale with a slider or command
✅ Developer-friendly - Focus on code, not infrastructure

Use Cases:

Rapid prototyping and MVPs
Startups and small teams
Side projects and demos
Development and staging environments
Low-traffic production applications

Heroku

One-Click Deploy

Manual Deployment

curl https://cli-assets.heroku.com/install.sh | sh

# Login to Heroku
heroku login

# Create app
heroku create tealtiger-app

# Set environment variables
heroku config:set OPENAI_API_KEY=sk-xxx
heroku config:set TEALTIGER_PROVIDER=openai
heroku config:set TEALTIGER_MODE=ENFORCE

# Deploy
git push heroku main

# Scale dynos
heroku ps:scale web=2

# View logs
heroku logs --tail

Procfile

# Procfile
web: gunicorn app:app --workers 4 --bind 0.0.0.0:$PORT

app.json (One-Click Deploy)

{
  "name": "TealTiger AI Security",
  "description": "AI agent security with TealTiger SDK",
  "repository": "https://github.com/agentguard-ai/tealtiger-heroku-template",
  "logo": "https://tealtiger.ai/logo.png",
  "keywords": ["ai", "security", "llm", "guardrails"],
  "env": {
    "OPENAI_API_KEY": {
      "description": "OpenAI API key",
      "required": true
    },
    "TEALTIGER_PROVIDER": {
      "description": "LLM provider (openai, anthropic, gemini, etc.)",
      "value": "openai",
      "required": true
    },
    "TEALTIGER_MODE": {
      "description": "Policy mode (ENFORCE, MONITOR, REPORT_ONLY)",
      "value": "ENFORCE",
      "required": true
    },
    "TEALTIGER_BUDGET_LIMIT": {
      "description": "Daily budget limit in USD",
      "value": "100",
      "required": false
    }
  },
  "formation": {
    "web": {
      "quantity": 1,
      "size": "basic"
    }
  },
  "addons": [
    {
      "plan": "papertrail:choklad",
      "as": "LOGGING"
    }
  ],
  "buildpacks": [
    {
      "url": "heroku/python"
    }
  ]
}

Python Example

# app.py
from flask import Flask, request, jsonify
from tealtiger import TealOpenAI, PolicyMode
import os

app = Flask(__name__)

# Initialize TealTiger client
client = TealOpenAI(
    api_key=os.environ['OPENAI_API_KEY'],
    mode=PolicyMode[os.environ.get('TEALTIGER_MODE', 'ENFORCE')],
    budget_limit=float(os.environ.get('TEALTIGER_BUDGET_LIMIT', 100))
)

@app.route('/chat', methods=['POST'])
def chat():
    data = request.json
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": data['message']}]
    )
    return jsonify({
        'response': response.choices[0].message.content
    })

if __name__ == '__main__':
    port = int(os.environ.get('PORT', 5000))
    app.run(host='0.0.0.0', port=port)

Railway

Deploy with Railway CLI

# Install Railway CLI
npm install -g @railway/cli

# Login
railway login

# Initialize project
railway init

# Set environment variables
railway variables set OPENAI_API_KEY=sk-xxx
railway variables set TEALTIGER_PROVIDER=openai
railway variables set TEALTIGER_MODE=ENFORCE

# Deploy
railway up

# View logs
railway logs

railway.json

{
  "$schema": "https://railway.app/railway.schema.json",
  "build": {
    "builder": "NIXPACKS"
  },
  "deploy": {
    "startCommand": "gunicorn app:app --workers 4 --bind 0.0.0.0:$PORT",
    "restartPolicyType": "ON_FAILURE",
    "restartPolicyMaxRetries": 10
  }
}

One-Click Deploy

Render

Deploy with render.yaml

# render.yaml
services:
  - type: web
    name: tealtiger-app
    env: python
    buildCommand: pip install -r requirements.txt
    startCommand: gunicorn app:app --workers 4 --bind 0.0.0.0:$PORT
    envVars:
      - key: OPENAI_API_KEY
        sync: false
      - key: TEALTIGER_PROVIDER
        value: openai
      - key: TEALTIGER_MODE
        value: ENFORCE
      - key: TEALTIGER_BUDGET_LIMIT
        value: "100"
    healthCheckPath: /health
    autoDeploy: true

Deploy from Dashboard

Go to render.com
Click “New +” → “Web Service”
Connect your Git repository
Configure:
- Name: tealtiger-app
- Environment: Python 3
- Build Command: pip install -r requirements.txt
- Start Command: gunicorn app:app --workers 4
Add environment variables
Click “Create Web Service”

One-Click Deploy

Fly.io

Deploy with Fly CLI

# Install Fly CLI
curl -L https://fly.io/install.sh | sh

# Login
fly auth login

# Launch app
fly launch

# Set secrets
fly secrets set OPENAI_API_KEY=sk-xxx
fly secrets set TEALTIGER_PROVIDER=openai
fly secrets set TEALTIGER_MODE=ENFORCE

# Deploy
fly deploy

# Scale
fly scale count 2

# View logs
fly logs

fly.toml

# fly.toml
app = "tealtiger-app"
primary_region = "iad"

[build]
  builder = "paketobuildpacks/builder:base"

[env]
  PORT = "8080"
  TEALTIGER_PROVIDER = "openai"
  TEALTIGER_MODE = "ENFORCE"

[http_service]
  internal_port = 8080
  force_https = true
  auto_stop_machines = true
  auto_start_machines = true
  min_machines_running = 1
  processes = ["app"]

[[services]]
  protocol = "tcp"
  internal_port = 8080
  processes = ["app"]

  [[services.ports]]
    port = 80
    handlers = ["http"]
    force_https = true

  [[services.ports]]
    port = 443
    handlers = ["tls", "http"]

  [services.concurrency]
    type = "connections"
    hard_limit = 25
    soft_limit = 20

  [[services.tcp_checks]]
    interval = "15s"
    timeout = "2s"
    grace_period = "1s"
    restart_limit = 0

  [[services.http_checks]]
    interval = "10s"
    timeout = "2s"
    grace_period = "5s"
    method = "get"
    path = "/health"
    protocol = "http"
    restart_limit = 0

[[vm]]
  cpu_kind = "shared"
  cpus = 1
  memory_mb = 256

Google App Engine

Deploy with gcloud

# Install gcloud CLI
curl https://sdk.cloud.google.com | bash

# Login
gcloud auth login

# Set project
gcloud config set project YOUR_PROJECT_ID

# Deploy
gcloud app deploy

# View logs
gcloud app logs tail -s default

app.yaml

# app.yaml
runtime: python311
entrypoint: gunicorn -b :$PORT app:app

env_variables:
  TEALTIGER_PROVIDER: "openai"
  TEALTIGER_MODE: "ENFORCE"
  TEALTIGER_BUDGET_LIMIT: "100"

automatic_scaling:
  target_cpu_utilization: 0.65
  min_instances: 1
  max_instances: 10
  min_pending_latency: 30ms
  max_pending_latency: automatic
  max_concurrent_requests: 50

handlers:
- url: /.*
  script: auto
  secure: always
  redirect_http_response_code: 301

Azure App Service

Deploy with Azure CLI

# Install Azure CLI
curl -sL https://aka.ms/InstallAzureCLIDeb | sudo bash

# Login
az login

# Create resource group
az group create --name tealtiger-rg --location eastus

# Create App Service plan
az appservice plan create \
  --name tealtiger-plan \
  --resource-group tealtiger-rg \
  --sku B1 \
  --is-linux

# Create web app
az webapp create \
  --resource-group tealtiger-rg \
  --plan tealtiger-plan \
  --name tealtiger-app \
  --runtime "PYTHON:3.11"

# Set environment variables
az webapp config appsettings set \
  --resource-group tealtiger-rg \
  --name tealtiger-app \
  --settings \
    OPENAI_API_KEY=sk-xxx \
    TEALTIGER_PROVIDER=openai \
    TEALTIGER_MODE=ENFORCE

# Deploy from Git
az webapp deployment source config \
  --name tealtiger-app \
  --resource-group tealtiger-rg \
  --repo-url https://github.com/your-org/tealtiger-app \
  --branch main \
  --manual-integration

Vercel (Edge Functions)

Deploy with Vercel CLI

# Install Vercel CLI
npm install -g vercel

# Login
vercel login

# Deploy
vercel

# Set environment variables
vercel env add OPENAI_API_KEY
vercel env add TEALTIGER_PROVIDER
vercel env add TEALTIGER_MODE

# Deploy to production
vercel --prod

vercel.json

{
  "version": 2,
  "builds": [
    {
      "src": "api/**/*.py",
      "use": "@vercel/python"
    }
  ],
  "routes": [
    {
      "src": "/api/(.*)",
      "dest": "/api/$1"
    }
  ],
  "env": {
    "TEALTIGER_PROVIDER": "openai",
    "TEALTIGER_MODE": "ENFORCE"
  }
}

Comparison Table

Platform	Deployment Time	Free Tier	Auto-Scaling	Custom Domains	Best For
Heroku	2-3 min	550 hrs/month	Yes	Yes	Rapid prototyping
Railway	1-2 min	$5 credit	Yes	Yes	Modern apps
Render	2-3 min	750 hrs/month	Yes	Yes	Production apps
Fly.io	1-2 min	3 VMs free	Yes	Yes	Global edge
App Engine	3-5 min	$300 credit	Yes	Yes	Google Cloud users
Azure App Service	3-5 min	60 min/day	Yes	Yes	Azure users
Vercel	1 min	100 GB-hrs	Yes	Yes	Edge functions

Cost Optimization

Tips for PaaS Deployments

Use auto-scaling to scale down during low traffic
Enable sleep mode for development environments
Use free tiers for staging and testing
Monitor usage with platform dashboards
Set budget alerts to avoid surprises

TealTiger Cost Tracking

from tealtiger import TealOpenAI, CostTracker

client = TealOpenAI(
    api_key=os.environ['OPENAI_API_KEY'],
    budget_limit=100,  # $100/day
    budget_window='daily'
)

# Get cost report
tracker = CostTracker(client)
report = tracker.get_daily_report()
print(f"Today's cost: ${report.total_cost:.2f}")

Monitoring

Health Check Endpoint

@app.route('/health')
def health():
    return jsonify({
        'status': 'healthy',
        'version': '1.1.0',
        'provider': os.environ.get('TEALTIGER_PROVIDER')
    })

Platform-Specific Monitoring

Heroku: Use Papertrail or Logentries add-ons
Railway: Built-in metrics dashboard
Render: Built-in metrics and logs
Fly.io: Built-in metrics and Prometheus integration
App Engine: Google Cloud Monitoring
Azure: Application Insights

Support

Documentation: docs.tealtiger.ai
GitHub Templates: github.com/agentguard-ai/tealtiger-templates
GitHub Issues: Report issues
Email: [email protected]

​Why PaaS?

​Heroku

​One-Click Deploy

​Manual Deployment

​Procfile

​app.json (One-Click Deploy)

​Python Example

​Railway

​Deploy with Railway CLI

​railway.json

​One-Click Deploy

​Render

​Deploy with render.yaml

​Deploy from Dashboard

​One-Click Deploy

​Fly.io

​Deploy with Fly CLI

​fly.toml

​Google App Engine

​Deploy with gcloud

​app.yaml

​Azure App Service

​Deploy with Azure CLI

​Vercel (Edge Functions)

​Deploy with Vercel CLI

​vercel.json

​Comparison Table

​Cost Optimization

​Tips for PaaS Deployments

​TealTiger Cost Tracking

​Monitoring

​Health Check Endpoint

​Platform-Specific Monitoring

​Support

Why PaaS?

Heroku

One-Click Deploy

Manual Deployment

Procfile

app.json (One-Click Deploy)

Python Example

Railway

Deploy with Railway CLI

railway.json

One-Click Deploy

Render

Deploy with render.yaml

Deploy from Dashboard

One-Click Deploy

Fly.io

Deploy with Fly CLI

fly.toml

Google App Engine

Deploy with gcloud

app.yaml

Azure App Service

Deploy with Azure CLI

Vercel (Edge Functions)

Deploy with Vercel CLI

vercel.json

Comparison Table

Cost Optimization

Tips for PaaS Deployments

TealTiger Cost Tracking

Monitoring

Health Check Endpoint

Platform-Specific Monitoring

Support