Amazon Bedrock

Connect your Amazon Bedrock-powered application to Latitude for observability.

Work in progress: Telemetry documentation is still being updated. Integration steps and APIs may be incomplete or out of date. Verify against your SDK versions and check back for revisions.

Overview

This guide shows you how to integrate Latitude Telemetry into an application that uses Amazon Bedrock.

You’ll keep calling Bedrock exactly as you do today. Telemetry simply observes and enriches those calls.

---

Requirements

• A Latitude account and API key
• A Latitude project slug
• A project that uses the AWS Bedrock SDK

---

Steps

1. Install

   TypeScript

   npm

   npm install @latitude-data/telemetry
   

   Python

   pip

   pip install latitude-telemetry
   
2. Initialize and use

   TypeScript

   import { initLatitude, capture } from "@latitude-data/telemetry"
   import {
     BedrockRuntimeClient,
     InvokeModelCommand,
   } from "@aws-sdk/client-bedrock-runtime"
   
   const latitude = initLatitude({
     apiKey: process.env.LATITUDE_API_KEY!,
     projectSlug: process.env.LATITUDE_PROJECT_SLUG!,
     instrumentations: ["bedrock"],
   })
   
   await latitude.ready
   
   const client = new BedrockRuntimeClient({ region: "us-east-1" })
   
   await capture("generate-reply", async () => {
     const command = new InvokeModelCommand({
       modelId: "anthropic.claude-3-haiku-20240307-v1:0",
       contentType: "application/json",
       body: JSON.stringify({
         anthropic_version: "bedrock-2023-05-31",
         max_tokens: 1024,
         messages: [{ role: "user", content: "Hello" }],
       }),
     })
     const response = await client.send(command)
     return JSON.parse(new TextDecoder().decode(response.body))
   })
   
   await latitude.shutdown()
   

   Python

   import json
   from latitude_telemetry import init_latitude, capture
   import boto3
   
   latitude = init_latitude(
       api_key="your-api-key",
       project_slug="your-project-slug",
       instrumentations=["bedrock"],
   )
   
   client = boto3.client("bedrock-runtime", region_name="us-east-1")
   
   def generate_reply():
       response = client.invoke_model(
           modelId="anthropic.claude-3-haiku-20240307-v1:0",
           contentType="application/json",
           body=json.dumps({
               "anthropic_version": "bedrock-2023-05-31",
               "max_tokens": 1024,
               "messages": [{"role": "user", "content": "Hello"}],
           }),
       )
       return json.loads(response["body"].read())
   
   capture("generate-reply", generate_reply)
   
   latitude.shutdown()
   

---

Streaming

When streaming, consume the stream inside capture() so the span covers the full operation:

TypeScript

import { InvokeModelWithResponseStreamCommand } from "@aws-sdk/client-bedrock-runtime"

await capture("stream-reply", async () => {
  const command = new InvokeModelWithResponseStreamCommand({
    modelId: "anthropic.claude-3-haiku-20240307-v1:0",
    contentType: "application/json",
    body: JSON.stringify({
      anthropic_version: "bedrock-2023-05-31",
      max_tokens: 1024,
      messages: [{ role: "user", content: input }],
    }),
  })
  const response = await client.send(command)

  for await (const event of response.body!) {
    if (event.chunk) {
      const data = JSON.parse(new TextDecoder().decode(event.chunk.bytes))
      if (data.type === "content_block_delta") {
        res.write(data.delta.text)
      }
    }
  }
  res.end()
})

Python

def stream_reply():
    response = client.invoke_model_with_response_stream(
        modelId="anthropic.claude-3-haiku-20240307-v1:0",
        contentType="application/json",
        body=json.dumps({
            "anthropic_version": "bedrock-2023-05-31",
            "max_tokens": 1024,
            "messages": [{"role": "user", "content": input}],
        }),
    )
    for event in response["body"]:
        chunk = json.loads(event["chunk"]["bytes"])
        if chunk.get("type") == "content_block_delta":
            yield chunk["delta"]["text"]

capture("stream-reply", stream_reply)

---

Seeing Your Traces

Once connected, traces appear automatically in Latitude:

1. Open your project in the Latitude dashboard
2. Each execution shows input/output messages, model, token usage, latency, and errors

---

That’s It

No changes to your Bedrock calls: just initialize Latitude and your LLM calls are traced.