This is documentation for Kamiwaza **1.0.0**, which is no longer actively maintained. For the current GA release, see [**1.0.1**](https://docs.kamiwaza.ai/).

Version: 1.0.0

## Creating a job

```python
from kamiwaza_sdk import KamiwazaClient

from kamiwaza_sdk.schemas.retrieval import RetrievalRequest

client = KamiwazaClient("https://localhost/api", api_key="...")

request = RetrievalRequest(

dataset_urn="urn:li:dataset:(urn:li:dataPlatform:s3,my-bucket/my-key,PROD)",

transport="inline",

format_hint="parquet",

credential_override='{"aws_access_key_id":"...","aws_secret_access_key":"..."}',

)

job = client.retrieval.create_job(request)

if job.inline:

print("Rows:", job.inline.data)

else:

print("Transport:", job.transport)
```

## Polling status

```python
status = client.retrieval.get_job(job.job_id)

print(status.status, status.progress)
```

## Streaming output

For SSE jobs (`transport="sse"`), `stream_job` yields raw server-sent-event lines:

```python
for event in client.retrieval.stream_job(job.job_id):

print(event)
```

> **Routing note:** the router is mounted under `/retrieval`, so the live paths
> are `/retrieval/jobs`, `/retrieval/jobs/{job_id}`, etc.

- [Creating a job](https://docs.kamiwaza.ai/sdk/1.0.0/current/services/retrieval#creating-a-job)
- [Polling status](https://docs.kamiwaza.ai/sdk/1.0.0/current/services/retrieval#polling-status)
- [Streaming output](https://docs.kamiwaza.ai/sdk/1.0.0/current/services/retrieval#streaming-output)