[All work](/work?type=work)

# ArticulateAI

ArticulateAI was one of my early voice experiments, recording a chosen amount of audio and passing it through transcription, a prompted response and optional speech output.

[Source code](https://github.com/Slaymish/ArticulateAI)

Independent development. Python, OpenAI and ElevenLabs APIs.

![Original ArticulateAI command menu, with twelve options for processing a spoken transcript.](https://cdn.sanity.io/images/qnuj1c4o/production/9e3bdfecff2b2c2c958917824aad775e98a5b9c4-1600x1000.png?w=1280\&q=80\&auto=format)

I was working with voice interfaces in 2023, and this version started with a recording length and a choice of prompt, so the same audio path could be used for a question, lecture notes or a recipe, with the selected prompt changing what happened to the transcript after the recording had finished.

The program recorded from the microphone, sent the audio to Whisper and saved the transcript with a timestamp, then asked GPT for a response using the selected instructions.

Short responses could go on to ElevenLabs for speech, though the implemented loop capped that at fewer than 100 characters, which kept this version close to a recording tool with spoken replies.

The catalogue date follows the first preserved commit on 17 May 2023.

## Outcome

An implemented audio-to-response workflow, including saved transcripts and short spoken replies. It used separate API calls, with the user choosing each recording window.

[All work](/work?type=work)