Skip to main content
  1. Sign up — Create an account at platform.respan.ai
  2. Create an API key — Generate one on the API keys page
  3. Add credits or a provider key — Add credits on the Credits page or connect your own provider key on the Integrations page

​
What is Haystack?

Haystack is a Python framework for building production-ready LLM pipelines with multiple components (retrievers, prompt builders, LLMs). The Respan integration captures your entire workflow execution and can route LLM calls through the Respan gateway.
Haystack tracing visualization

​
Setup

1

Install packages

pip install respan-exporter-haystack
2

Set environment variables

.env
RESPAN_API_KEY=your-respan-api-key
OPENAI_API_KEY=your-openai-key
HAYSTACK_CONTENT_TRACING_ENABLED=true
The HAYSTACK_CONTENT_TRACING_ENABLED variable activates Haystack’s tracing system.
3

Add RespanConnector to your pipeline

import os
from haystack import Pipeline
from haystack.components.builders import PromptBuilder
from haystack.components.generators import OpenAIGenerator
from respan_exporter_haystack import RespanConnector

os.environ["HAYSTACK_CONTENT_TRACING_ENABLED"] = "true"

pipeline = Pipeline()
pipeline.add_component("tracer", RespanConnector("My Workflow"))
pipeline.add_component("prompt", PromptBuilder(template="Tell me about {{topic}}."))
pipeline.add_component("llm", OpenAIGenerator(model="gpt-4o-mini"))
pipeline.connect("prompt", "llm")

result = pipeline.run({"prompt": {"topic": "artificial intelligence"}})
print(result["llm"]["replies"][0])
print(f"\nTrace URL: {result['tracer']['trace_url']}")
4

View your trace

After running, you’ll get a trace URL. Visit it to see the pipeline execution timeline, each component’s input/output, timing, and token usage.Dashboard: platform.respan.ai/platform/traces

​
Gateway

Route LLM calls through the Respan gateway for automatic logging, fallbacks, and cost optimization. Replace OpenAIGenerator with RespanGenerator.
Haystack gateway integration

​
Basic usage

import os
from haystack import Pipeline
from haystack.components.builders import PromptBuilder
from respan_exporter_haystack import RespanGenerator

pipeline = Pipeline()
pipeline.add_component("prompt", PromptBuilder(template="Tell me about {{topic}}."))
pipeline.add_component("llm", RespanGenerator(
    model="gpt-4o-mini",
    api_key=os.getenv("RESPAN_API_KEY")
))
pipeline.connect("prompt", "llm")

result = pipeline.run({"prompt": {"topic": "machine learning"}})
print(result["llm"]["replies"][0])

​
Attributes

Pass Respan-specific parameters via generation_kwargs on RespanGenerator: 
from respan_exporter_haystack import RespanGenerator

pipeline.add_component("llm", RespanGenerator(
    model="gpt-4o-mini",
    api_key=os.getenv("RESPAN_API_KEY"),
    generation_kwargs={
        "customer_identifier": "user_123",
        "thread_identifier": "conversation_456",
        "metadata": {"session_id": "abc123"},
    }
))
AttributeDescription
customer_identifierCustomer or user identifier
thread_identifierThread or conversation identifier
metadataCustom key-value pairs attached to the trace
fallback_modelsList of fallback models (gateway feature)

​
Prompts

Use platform-managed prompts for centralized control: 
import os
from haystack import Pipeline
from respan_exporter_haystack import RespanGenerator

pipeline = Pipeline()
pipeline.add_component("llm", RespanGenerator(
    prompt_id="your-prompt-id",
    api_key=os.getenv("RESPAN_API_KEY")
))

result = pipeline.run({
    "llm": {
        "prompt_variables": {
            "user_input": "your text here"
        }
    }
})
Create prompts at: platform.respan.ai/platform/prompts

​
Observability

With this integration, Respan auto-captures:
  • Pipeline execution — the full pipeline as a trace
  • Component calls — each component’s input/output as a span
  • LLM calls — model, token usage, timing
  • Gateway features — fallbacks, load balancing, cost tracking
  • Errors — failed components and error details
View traces on the Traces page.