Updated Jun 15, 2026 · by AutoCoder · 2 min readAI-assisted
Last updated
Poll a submitted coding task by its id until the state is Completed.
| Field | Value |
|---|---|
| Endpoint | https://services.hank.ai/autocoding/v1/tasks/<id> |
| Method | GET |
| API key header | x-api-key |
| State | Meaning |
|---|---|
In-Progress | The task is still coding. Poll again after a short delay. |
Completed | The result is ready in response.result. |
Error | The task failed. Inspect the error field and contact support if it persists. |
Poll at a modest interval, for example every 5 to 15 seconds. A typical case completes within a few minutes.
{
"id": "<id>",
"response": {
"result": {
"job_id": "<your internal id from the POST, if supplied>",
"entities": [
{
"entity": {
"id": "<unique-within-request id>",
"provider": "<provider name>",
"entityValue": [
{
"id": "<unique-within-request id>",
"code": 20610,
"confidence": 0.98,
"noteType": "SurgeonProcedureNote",
"noteId": 25
}
],
"order": 1,
"Line_item_number": 1,
"type": "SurgCPT"
}
}
]
}
},
"state": "Completed",
"status": null,
"metadata": {
"apiKey": "<your-api-key>",
"timeStamp": "timestamp",
"processStartTime": "time",
"processEndTime": "time"
},
"error": null
}Each entry in entities is one coded line item:
| Field | Notes |
|---|---|
entityValue[] | The codes of this line item's type, in order. Each carries the code, its confidence (0 to 1), and the noteType + noteId of the source document. |
type | The line-item code type: SurgCPT, AnesCPT, ICD10, Modifier, or Practitioner. |
order | The position of this line item within its type. |
Line_item_number | The claim line number. |
provider | The provider the line item belongs to. |
Use confidence to route low-confidence codes into your own review workflow. Match results back to your system with job_id (from your POST) and the noteId references.