Version 1.1 — AI Teaching Summaries
Expand the existing transcription workflow by using the OpenAI API to automatically convert completed transcripts into structured teaching notes.
What It Does
- Reads an existing `.txt` transcript.
- Reads instructions from a separate Markdown prompt file.
- Combines the prompt with the transcript.
- Sends the combined text to GPT through the OpenAI API.
- Receives a structured teaching summary.
- Saves the summary as a Markdown `.md` file.
- Skips summaries that already exist.
- Uses `try/except` to handle AI summarization errors.
- Tracks summaries as created, skipped, or failed.
- Includes summary results in the final processing report.
New Concepts I Learned
- API calls - OpenAI client
- Environment variables for API keys
- Reading Markdown files with Python
- Combining multiple strings into a complete AI prompt
- Using descriptive return values:
"created"`
"skipped"`
"failed"`
- Using a function's return value to update counters in `main()`
- Separating AI instructions from Python logic with a dedicated prompt file
Completed Workflow
Raw Video → Faster Whisper → `.txt` transcript → Custom prompt → GPT → Structured `.md` teaching summary