---
title: Exa
description: Instrument Exa semantic search calls and send traces to OpenObserve via OpenTelemetry.
---

# **Exa → OpenObserve**

Capture latency, query text, and result counts for every Exa semantic search call in your AI pipeline. Exa does not have a dedicated OTel instrumentor, so instrumentation uses manual OpenTelemetry spans that wrap each search request.

## **Prerequisites**

* Python 3.8+
* An [OpenObserve](https://openobserve.ai/) account (cloud or self-hosted)
* Your OpenObserve **organisation ID** and **Base64-encoded auth token**
* An [Exa](https://exa.ai/) API key

## **Installation**

```shell
pip install openobserve-telemetry-sdk exa-py opentelemetry-api python-dotenv
```

## **Configuration**

Create a `.env` file in your project root:

```
OPENOBSERVE_URL=https://api.openobserve.ai/
OPENOBSERVE_ORG=your_org_id
OPENOBSERVE_AUTH_TOKEN=Basic <your_base64_token>
EXA_API_KEY=your-exa-api-key
```

## **Instrumentation**

Call `openobserve_init()` **before** importing Exa. Wrap each search call in a `tracer.start_as_current_span()` block and record query attributes and result counts.

```python
from dotenv import load_dotenv
load_dotenv()

from openobserve import openobserve_init
openobserve_init()

from opentelemetry import trace
import os
from exa_py import Exa

tracer = trace.get_tracer(__name__)
client = Exa(api_key=os.environ["EXA_API_KEY"])

def search(query: str, num_results: int = 5):
    with tracer.start_as_current_span("exa.search") as span:
        span.set_attribute("exa.query", query)
        span.set_attribute("exa.num_results", num_results)
        result = client.search(query, num_results=num_results)
        span.set_attribute("exa.result_count", len(result.results))
        return result

result = search("OpenTelemetry distributed tracing")
for r in result.results:
    print(r.title, r.url)
```

## **What Gets Captured**

| Attribute | Description |
| ----- | ----- |
| `exa_query` | The semantic search query string |
| `exa_num_results` | Number of results requested |
| `exa_result_count` | Number of results returned |
| `span_status` | `OK` or error status |
| `duration` | End-to-end search latency |

## **Viewing Traces**

1. Log in to OpenObserve and navigate to **Traces**
2. Filter by span name `exa.search` to see all search calls
3. Click a span to inspect the query and result count
4. Sort by duration to find the slowest searches

![Exa search trace in OpenObserve](../../../images/integration/ai/exa.png)

## **Next Steps**

With Exa instrumented, every search call in your RAG or agent pipeline is recorded in OpenObserve. From here you can track search latency, monitor result quality over time, and correlate search spans with downstream LLM spans.

## **Read More**

- [LLM Observability Overview](../llm-applications.md)
- [Traces Ingestion with Python](../../../ingestion/traces/python.md)
- [Exploring Traces in OpenObserve](../../../user-guide/data-exploration/traces/index.md)
- [Building Dashboards](../../../user-guide/analytics/dashboards/index.md)