> ## Documentation Index
> Fetch the complete documentation index at: https://docs.portkey.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Deploy Portkey in Your Infrastructure

> Get your AI gateway running in 4 simple steps. Most teams finish in under 2 hours.

<div className="hero-section">
  # Your AI Gateway. Your Infrastructure. **2 Hours.**

  Deploy Portkey's data plane in your VPC and start routing LLM traffic securely. No data leaves your network.

  <div className="success-metrics">
    <div className="metric">
      <span className="number">247</span>
      <span className="label">Teams deployed</span>
    </div>

    <div className="metric">
      <span className="number">2.3 hrs</span>
      <span className="label">Average time</span>
    </div>

    <div className="metric">
      <span className="number">15 min</span>
      <span className="label">To first API call</span>
    </div>
  </div>
</div>

***

## 🎯 What You'll Have in 2 Hours

<div className="outcome-grid">
  <Card icon="rocket" title="Working AI Gateway">
    Route requests to OpenAI, Anthropic, and 1600+ LLMs through your private endpoint
  </Card>

  <Card icon="shield" title="Complete Data Privacy">
    All prompts and responses stay in your VPC. Only metrics leave (no sensitive data).
  </Card>

  <Card icon="chart-line" title="Full Observability">
    See every request, cost, latency, and error in the Portkey dashboard
  </Card>
</div>

<Frame>
  <img src="https://mintlify.s3.us-west-1.amazonaws.com/portkey-docs/images/deployment-success.png" alt="Successful deployment showing gateway handling requests" />
</Frame>

***

## 📍 Your 4-Step Journey

<Steps>
  <Step title="Check Prerequisites" duration="5 min" icon="clipboard-check">
    Verify your Kubernetes cluster and create storage buckets
  </Step>

  <Step title="Get Credentials" duration="10 min" icon="key">
    Receive and test your access tokens from Portkey
  </Step>

  <Step title="Deploy Gateway" duration="30 min" icon="rocket">
    Run one Helm command with our production-ready template
  </Step>

  <Step title="Validate Success" duration="15 min" icon="check-circle">
    Make your first API call and see logs flowing
  </Step>
</Steps>

<Alert type="success" icon="phone">
  **We're on standby**: Our deployment team typically responds in \<15 minutes during business hours. Don't hesitate to ping us!
</Alert>

***

## Step 1: Check Prerequisites (5 min)

<div className="checklist-card">
  ### ✓ Quick Checks

  ```bash theme={"system"}
  # 1. Kubernetes ready?
  kubectl version --short
  # Need: v1.24+

  # 2. Helm installed?
  helm version --short
  # Need: v3.0+

  # 3. Can reach internet?
  curl -I https://control.portkey.ai
  # Need: HTTP/2 200
  ```

  ### ✓ Pick Your Storage

  <Tabs>
    <Tab title="AWS (S3)" icon="aws">
      ```bash theme={"system"}
      # Create a bucket
      aws s3 mb s3://my-portkey-logs
      ```
    </Tab>

    <Tab title="GCS" icon="google">
      ```bash theme={"system"}
      # Create a bucket
      gsutil mb gs://my-portkey-logs
      ```
    </Tab>

    <Tab title="Quick Test (Wasabi)" icon="bolt">
      We'll provide temporary credentials - no setup needed!
    </Tab>
  </Tabs>
</div>

<Card title="Need help? Start a quick chat" icon="message" href="https://portkey.wiki/chat">
  Most issues resolved in under 10 minutes
</Card>

***

## Step 2: Get Credentials (10 min)

You'll receive a **1Password link** with two items:

<div className="credential-cards">
  <div className="cred-card">
    <h4>🐳 Docker Access</h4>
    <p>Username & password for our container registry</p>
  </div>

  <div className="cred-card">
    <h4>🔐 Gateway Token</h4>
    <p>JWT to connect your gateway to control plane</p>
  </div>
</div>

### Quick Test

```bash theme={"system"}
# 1. Test Docker access
docker login -u <username> -p <password>
docker pull portkey/gateway-enterprise:latest

# 2. Create Kubernetes secret
kubectl create namespace portkey
kubectl -n portkey create secret docker-registry portkey-creds \
  --docker-username=<username> \
  --docker-password=<password>
```

<Alert type="warning">
  **Can't pull image?** It's usually a firewall. [Click here for instant help →](https://portkey.wiki/slack)
</Alert>

***

## Step 3: Deploy Gateway (30 min)

### 3.1 Save This Config

<div className="config-wizard">
  Create `values.yaml` with your details:

  ```yaml theme={"system"}
  # values.yaml - Production-ready config
  replicaCount: 2

  images:
    gatewayImage:
      repository: "portkey/gateway-enterprise"
      tag: "1.10.24"

  environment:
    data:
      # Your identifiers
      SERVICE_NAME: "my-company-gateway"
      PORTKEY_CLIENT_AUTH: "<token-from-1password>"

      # Storage (pick one)
      LOG_STORE: "s3"  # or "gcs", "mongo", "wasabi"
      LOG_STORE_REGION: "us-east-1"
      LOG_STORE_GENERATIONS_BUCKET: "my-portkey-logs"
      LOG_STORE_ACCESS_KEY: "<your-access-key>"
      LOG_STORE_SECRET_KEY: "<your-secret-key>"

      # Cache (auto-deployed)
      CACHE_STORE: "redis"
      REDIS_URL: "redis://portkey-redis:6379"

      # Analytics
      ANALYTICS_STORE: "control_plane"

  # Expose your gateway
  ingress:
    enabled: true
    className: "nginx"
    hosts:
      - host: gateway.internal.mycompany.com
        paths:
          - path: /
            pathType: Prefix

  # Auto-scaling
  autoscaling:
    enabled: true
    minReplicas: 2
    maxReplicas: 10
  ```
</div>

### 3.2 Deploy with One Command

```bash theme={"system"}
# Add Portkey charts
helm repo add portkey https://portkey-ai.github.io/helm
helm repo update

# Deploy!
helm upgrade --install portkey-gateway portkey/gateway \
  -n portkey -f values.yaml

# Watch it come up
kubectl -n portkey get pods -w
```

<div className="deploy-status">
  Expected output:

  ```
  NAME                                READY   STATUS
  portkey-gateway-7f9b8c5d4-abc123   1/1     Running
  portkey-gateway-7f9b8c5d4-def456   1/1     Running
  portkey-redis-master-0             1/1     Running
  ```
</div>

***

## Step 4: Validate Success (15 min)

### 4.1 Check Health

```bash theme={"system"}
# Port-forward for quick test
kubectl -n portkey port-forward svc/portkey-gateway 8787:8787

# In new terminal
curl http://localhost:8787/v1/health
# ✅ Should return: {"status":"healthy"}
```

### 4.2 Get Your API Keys

<div className="key-steps">
  1. Go to [app.portkey.ai/model-catalog](https://app.portkey.ai/model-catalog)
  2. Click **"Add Provider"**
  3. Select your LLM provider (e.g., OpenAI)
  4. Enter your OpenAI API key
  5. Name your provider (e.g., `openai-prod`)
</div>

### 4.3 Make Your First Call! 🎉

```bash theme={"system"}
# Using your gateway URL
curl https://gateway.internal.mycompany.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "x-portkey-api-key: <your-portkey-api-key>" \
  -H "x-portkey-provider: @openai-prod" \
  -d '{
    "model": "gpt-4",
    "messages": [{"role": "user", "content": "Hello from my private gateway!"}]
  }'
```

### 4.4 See It in Action

<div className="validation-grid">
  <Card title="✅ Check Logs" icon="file-text">
    Visit app.portkey.ai → Logs
    Your request appears instantly
  </Card>

  <Card title="✅ View Analytics" icon="chart-bar">
    Visit app.portkey.ai → Analytics
    See costs and latencies
  </Card>
</div>

<Alert type="success" icon="trophy">
  **Congratulations!** Your AI gateway is live. All LLM traffic now routes through your infrastructure.
</Alert>

***

## 🎊 You Did It! What's Next?

<div className="next-steps-grid">
  <Card title="Add More Providers" icon="plus-circle" href="/docs/integrations/llms">
    Connect Claude, Gemini, Llama, and 1600+ models
  </Card>

  <Card title="Set Up Fallbacks" icon="shield-alt" href="/docs/product/ai-gateway/fallbacks">
    Add redundancy with automatic provider failover
  </Card>

  <Card title="Configure Caching" icon="lightning" href="/docs/product/ai-gateway/cache-simple-and-semantic">
    Reduce costs by 40% with semantic caching
  </Card>

  <Card title="Production Checklist" icon="clipboard-list" href="/docs/self-hosting/hybrid-deployments/architecture">
    Security hardening and monitoring setup
  </Card>
</div>

<Card title="📞 Book Your Success Review" icon="calendar-check" href="https://portkey.wiki/success-review">
  Optional: Show us your deployment and get optimization tips (15 min)
</Card>

***

## 🆘 Troubleshooting

<details>
  <summary>Common Issues (Click to expand)</summary>

  <Accordion title="Pods won't start">
    ```bash theme={"system"}
    # Check logs
    kubectl -n portkey logs -l app=portkey-gateway

    # Common fixes:
    - Verify credentials in 1Password
    - Check firewall allows Docker Hub
    - Ensure namespace has image pull secret
    ```
  </Accordion>

  <Accordion title="Can't reach gateway">
    ```bash theme={"system"}
    # Test internal service
    kubectl -n portkey port-forward svc/portkey-gateway 8787:8787
    curl http://localhost:8787/v1/health

    # Check ingress
    kubectl -n portkey describe ingress
    ```
  </Accordion>

  <Accordion title="No logs appearing">
    * Verify S3/storage credentials
    * Check gateway can reach storage endpoint
    * Look for errors in: `kubectl -n portkey logs -l app=portkey-gateway | grep ERROR`
  </Accordion>
</details>

***

## 💬 Get Help Fast

<Card title="Live Chat" icon="message-circle">
  [portkey.wiki/chat](https://portkey.wiki/chat)
  \~10 min response time
</Card>

<Card title="Slack Community" icon="slack">
  [portkey.wiki/slack](https://portkey.wiki/slack)
  \#deployment-help channel
</Card>

<Card title="Email Support" icon="mail">
  [support@portkey.ai](mailto:support@portkey.ai)
  24-hour response SLA
</Card>

***

<Card title="Portkey is now PRISMA AIRS AI Gateway. See it in action." href="https://www.paloaltonetworks.in/ai-security/ai-gateway?utm_source=portkey&utm_medium=referral&utm_campaign=prisma_airs&utm_content=docs_nav#contact" icon="arrow-up-right-from-square">
  Contact Us
</Card>