Freshman year's 1st ChatGPT API project & 1st hackathon project. Utilized GPT3.5 LLM to perform summarizing (recapping), sentiment analysis, keypoints extraction, action items extraction of an audio file and save AI responses into a word docx.
Run .ipynb file and EarningCalls.wav's full "recap" should be created in your local folder as a file name "recap.docx".
v1.0.0: Due to the short 2-day time limit at TigerHacks2023, the program could only create a recap for 1 specific audio file. Definitely more improvements necessary. *OpenAI has since updated its API, thus, I will also make code changes accordingly
v2.0.0 update: Can now upload any audio files and they can be of longer lengths! Can also choose whether the recap should be in pdf, docx, and/or txt format(s).
Done: This independent project will continue to be updated upon by me. I plan to improve the project so that audio files of any length can be "recapped".
This project is mainly derived from ChatGPT API's tutorial: https://platform.openai.com/docs/tutorials/meeting-minutes
Additional sources used are listed in 'sources used.txt' file.