Haystack | Respan Docs

Haystack is a Python framework for building production-ready LLM pipelines. It provides components for retrieval, generation, and processing that can be composed into pipelines. Respan gives you full observability over every pipeline run, component, and LLM call — and gateway routing through the OpenAI-compatible Respan endpoint.

Set up Respan

Create an account at platform.respan.ai and grab an API key. For gateway, also add credits or a provider key.

Run npx @respan/cli setup to set up with your coding agent.

Example projects

Python examples

Tracing

Gateway

Setup

Install packages

$ pip install respan-exporter-haystack haystack-ai

Set environment variables

$ export OPENAI_API_KEY="YOUR_OPENAI_API_KEY"
$ export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"

OPENAI_API_KEY is used for LLM requests. RESPAN_API_KEY is used to export traces to Respan.

Add RespanConnector to your pipeline

1 import os
2 from haystack import Pipeline
3 from haystack.components.generators import OpenAIGenerator
4 from respan_exporter_haystack import RespanConnector
5 
6 pipeline = Pipeline()
7 pipeline.add_component("respan", RespanConnector(api_key=os.environ["RESPAN_API_KEY"]))
8 pipeline.add_component("llm", OpenAIGenerator(model="gpt-4.1-nano"))
9 pipeline.connect("respan", "llm")
10 
11 result = pipeline.run({"respan": {"prompt": "Tell me a joke about AI"}})
12 print(result)

View your trace

Open the Traces page to see your pipeline run with component spans, LLM calls, and inputs/outputs.

Configuration

See the Haystack Exporter SDK reference for the full API.

Component	Description
`RespanConnector`	Connects pipelines to Respan for tracing.
`RespanTracer`	Alternative tracing integration.
`RespanGenerator`	Routes LLM calls through the Respan gateway.
`RespanChatGenerator`	Chat-specific gateway component.

Attributes

Pass Respan attributes through the connector.

1 result = pipeline.run({
2     "respan": {
3         "prompt": "Tell me a joke",
4         "customer_identifier": "user-123",
5         "metadata": {"team": "ml"},
6     }
7 })

Attribute	Description
`customer_identifier`	User/customer identifier.
`metadata`	Custom key-value pairs.

Prompts

Use Respan-managed prompts in Haystack pipelines.

1 from respan_exporter_haystack import RespanGenerator
2 
3 generator = RespanGenerator(
4     api_key=os.environ["RESPAN_API_KEY"],
5     model="gpt-4.1-nano",
6     prompt_id="your-prompt-id",
7     prompt_version=1,
8 )