Deepgram Deploy Integration

Table of Contents

• Overview
• Prerequisites
• Instructions
• Output
• Error Handling
• Examples
• Resources

Overview

Deploy Deepgram integrations to various cloud platforms and container environments including Docker, Kubernetes, AWS Lambda, Google Cloud Run, and Vercel.

Prerequisites

• Production API key ready
• Infrastructure access configured
• Secret management in place
• Monitoring configured

Instructions

Step 1: Containerize with Docker

Build a multi-stage Dockerfile with non-root user, health checks, and production optimizations.

Step 2: Configure Orchestration

Choose deployment target: Docker Compose (simple), Kubernetes (scale), or serverless (event-driven).

Step 3: Manage Secrets

Use platform-native secret management (K8s Secrets, AWS Secrets Manager, GCP Secret Manager).

Step 4: Set Up Health Checks

Configure liveness and readiness probes against /health endpoint.

Step 5: Configure Auto-scaling

Set up HPA (Kubernetes) or managed scaling with CPU-based thresholds at 70% utilization.

Step 6: Create Deploy Script

Build environment-aware deploy script that runs build, tests, deploys, and smoke tests.

Output

• Dockerfile with multi-stage build and health checks
• Docker Compose or Kubernetes manifests
• Serverless configuration (Lambda/Cloud Run/Vercel)
• Deploy script with environment selection
• Auto-scaling configuration

Error Handling

Issue	Cause	Resolution
Container fails to start	Missing secret or env var	Check secret mounts and env config
Health check failing	Service not ready	Increase initialDelaySeconds
OOM kills	Memory limit too low	Increase memory limit (512Mi minimum)
Connection refused	Wrong port mapping	Verify containerPort matches app

Examples

Quick Docker Build & Run

set -euo pipefail
docker build -t deepgram-service .
docker run -p 3000:3000 -e DEEPGRAM_API_KEY=$DEEPGRAM_API_KEY deepgram-service  # 3000: 3 seconds in ms

Deploy Script Usage

./scripts/deploy.sh staging    # Deploy to staging
./scripts/deploy.sh production # Deploy to production

See detailed implementation for advanced patterns.

Resources

• Docker Best Practices
• Kubernetes Deployments
• AWS Lambda with Node.js
• Google Cloud Run