Mini Challenge 1: Serverless App

Goal: create a serverless app that can 1/ send a short audio file to OpenAI Whisper API for transcription 2/ store OpenAI response in a text file and put it in S3

Here are the instructions given by ChatGPT 🙂

1. Set up AWS Services:
   • Create an AWS Lambda function: This function will execute your code in response to an event, such as an audio file being uploaded.
   • Create an S3 bucket: This bucket will be used to store both the input audio files and the resulting text files.
2. Configure AWS Lambda Environment:
   • Set up the Lambda function environment with the necessary dependencies and runtime. You may need to include the OpenAI API key and other configurations as environment variables.
3. Write Lambda Function Code:
   • Use a programming language supported by AWS Lambda (e.g., Python, Node.js).
   • Use an S3 trigger to execute the Lambda function when a new audio file is uploaded.
   • In the Lambda function code, use the OpenAI API to transcribe the audio file.
   • Store the OpenAI response in a text file.
4. Store Text File in S3:
   • Upload the text file to the same or a different S3 bucket.
   • Ensure appropriate permissions for the Lambda function to interact with the S3 bucket.
5. Test the Setup:
   • Upload a short audio file to the S3 bucket to trigger the Lambda function.
   • Verify that the transcription is performed, and the resulting text file is stored in the S3 bucket.
6. Monitoring and Error Handling:
   • Implement logging within your Lambda function to capture errors and monitor execution.
   • Set up appropriate error handling mechanisms to ensure the robustness of your serverless app.

Seems pretty reasonable to me, so I decide to give it a try.