Langfuse Deploy Integration

Langfuse Deploy Integration

Overview

Deploy Langfuse LLM observability with your application. Covers integrating the Langfuse SDK for tracing, deploying self-hosted Langfuse via Docker, and configuring cloud-hosted Langfuse across Vercel, Docker, and Cloud Run deployments.

Prerequisites

• Langfuse account (cloud at langfuse.com or self-hosted)
• Langfuse public/secret key pair
• Application using langfuse SDK
• Platform CLI installed

Instructions

Step 1: Configure Langfuse SDK

// lib/langfuse.ts
import { Langfuse } from "langfuse";

export const langfuse = new Langfuse({
  publicKey: process.env.LANGFUSE_PUBLIC_KEY!,
  secretKey: process.env.LANGFUSE_SECRET_KEY!,
  baseUrl: process.env.LANGFUSE_HOST || "https://cloud.langfuse.com",
});

// Ensure traces are flushed before serverless function exits
export async function flushLangfuse() {
  await langfuse.flushAsync();
}

Step 2: Vercel Deployment

vercel env add LANGFUSE_PUBLIC_KEY production
vercel env add LANGFUSE_SECRET_KEY production
vercel env add LANGFUSE_HOST production  # Optional for self-hosted
vercel --prod

// api/chat.ts - Trace LLM calls
import { langfuse, flushLangfuse } from "../lib/langfuse";

export async function POST(req: Request) {
  const trace = langfuse.trace({ name: "chat-request" });

  const generation = trace.generation({
    name: "llm-call",
    model: "gpt-4o",
    input: messages,
  });

  const response = await openai.chat.completions.create({ model: "gpt-4o", messages });

  generation.end({ output: response.choices[0].message });
  await flushLangfuse();

  return Response.json(response);
}

Step 3: Self-Hosted Langfuse (Docker)

# docker-compose.yml
version: "3.8"
services:
  langfuse:
    image: langfuse/langfuse:latest
    ports:
      - "3001:3000"  # 3000: 3001: alternate dev port
    environment:
      - DATABASE_URL=postgresql://user:pass@postgres:5432/langfuse  # 5432: PostgreSQL port
      - NEXTAUTH_SECRET=your-secret
      - NEXTAUTH_URL=http://localhost:3001  # alternate dev port
      - SALT=your-salt
    depends_on:
      - postgres

  postgres:
    image: postgres:16
    environment:
      - POSTGRES_USER=user
      - POSTGRES_PASSWORD=pass
      - POSTGRES_DB=langfuse
    volumes:
      - pgdata:/var/lib/postgresql/data

volumes:
  pgdata:

Step 4: LangChain Integration

from langfuse.callback import CallbackHandler

langfuse_handler = CallbackHandler(
    public_key=os.environ["LANGFUSE_PUBLIC_KEY"],
    secret_key=os.environ["LANGFUSE_SECRET_KEY"],
    host=os.environ.get("LANGFUSE_HOST", "https://cloud.langfuse.com"),
)

# Attach to any LangChain chain
chain.invoke({"input": "Hello"}, config={"callbacks": [langfuse_handler]})

Step 5: Health Check

export async function GET() {
  try {
    await langfuse.flushAsync();
    return Response.json({ status: "healthy", tracing: "enabled" });
  } catch {
    return Response.json({ status: "degraded", tracing: "disabled" }, { status: 503 });  # HTTP 503 Service Unavailable
  }
}

Error Handling

Examples

Quick Integration Test

const trace = langfuse.trace({ name: "test" });
trace.update({ output: "test complete" });
await langfuse.flushAsync();

Resources

• Langfuse Documentation
• Langfuse Self-Hosting
• Langfuse SDK

Next Steps

For webhook handling, see langfuse-webhooks-events.

Output

• Configuration files or code changes applied to the project
• Validation report confirming correct implementation
• Summary of changes made and their rationale