> ## Documentation Index
> Fetch the complete documentation index at: https://docs.corti.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Submit Feedback

The job of improving speech-to-text accuracy, coverage, and performance is never complete. That's why collecting real world feedback and input is essential for continuous improvement in the areas that matter most.

Use the form below to share any detailed feedback or specific examples. The feedback will be used to analyze where model or post-processing improvements can be made in future releases.

***

## Feedback details

Issues can be classified as one of the following types:

* **Recognition** - speech to text output was different than expected
* **Formatting** - speech to text output was correct, but not in the style as expected
* **Command** - command was not recognized or returned as text instead of command
* **Spelling** - speech to text output was correct, but with a spelling error
* **Enhancement request** - desired new feature or functionality

For optimal support, please include a completed <a href="https://drive.google.com/uc?export=download&id=1oxYzHap59hERqIjDznPQ-klUbVT8nA-b" download>STT feedback details csv</a> with your submission. It includes the following fields:

| Field                   | Required    | Description                                                                                                                                                                  |
| ----------------------- | ----------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `term`                  | Required    | Expected written form in transcript output. May include spaces, digits, and special characters (e.g. "CO2").                                                                 |
| `spokenForm`            | Recommended | The normalized form of the term. Initialisms should be separated by hyphens; acronyms can be represented without hyphenation; digits should be written out (e.g. "C-O-two"). |
| `misrecognizedOutput`   | Optional    | Incorrect speech to text output that was received for the defined term (e.g., "see O2").                                                                                     |
| `sampleAudioPath`       | Optional    | The filename for a provided audio file.                                                                                                                                      |
| `sampleAudioTranscript` | Optional    | Expected text output for a provided audio file.                                                                                                                              |

***

## Feedback form

Prefer a full-page view? [Click here](https://corti.fillout.com/t/8DwE3qbbXRus) to open the form in a new tab.

<div className="mt-6 overflow-hidden rounded-2xl border border-slate-200 bg-white shadow-sm">
  <iframe src="https://corti.fillout.com/t/8DwE3qbbXRus" title="Speech-to-text feedback form" style={{ width: '100%', height: '1000px', minHeight: '620px' }} />
</div>

<br />

<Note>Need further support? [Contact us](mailto:help@corti.ai) directly!</Note>