Question 1

Is this a qualitative analysis or research tool?

Accepted Answer

No. Review, coding, and insight work live in a **different layer** — often in tools built for that purpose. This project is the **pipeline**: how to reliably get a recording to a transcript in your flow, with status and correction. Text becomes organisational data you can search, export, or hand to a separate analysis step if you need one.

Question 2

Do you build your own speech recognition?

Accepted Answer

No. We integrate external STT APIs. Our work is storage, queue, worker, webhooks, and the recording record in the database — so changing provider does not mean rebuilding everything.

Question 3

Why webhooks instead of polling every few seconds?

Accepted Answer

STT takes long relative to an HTTP request. The webhook updates the record when the provider finishes — without making users wait in the browser.

Question 4

Must the user wait on upload?

Accepted Answer

Not for transcription. After the file is stored and the record created, the rest runs in the background. The team returns when status shows readiness for correction.

Question 5

Can transcripts be corrected before use?

Accepted Answer

Yes — that is part of the flow. STT supplies a draft; the organisation approves text before it becomes documentation or material for further analysis.

Question 6

How long does delivery take?

Accepted Answer

It depends on volume and quality requirements. We start with one recording type and one upload → STT → persist path, then extend — without unrealistic dates.

Call recordings are organisational data — without a flow they never become durable knowledge