top of page



Nazanin Komeilizadeh, Roberto Nunez, Korel Gundem, Diliya Yalikun, Aryama Singh


Create an AI assisted resource (a conversational bot) that efficiently collates finance and economy related current news within a specified time period (eg. 1 week) from a YouTube news channel (eg. Bloomberg) where people can ask about finance related current news that simultaneously shows the sentiment associated with the context of the given query.
The knowledge base for this would be a YouTube News channels (eg. Bloomberg) generated within 1 week:
The core technologies used in this project would be :
1. pytube library to download YT videos and generate audio.
2. Whisper openAi for Audio-to-Text conversion
3. Transcribing + Refining of text
4. Vectorization of YT videos- Creating a vectorized representation of the transcribed text from YouTube videos for storage and retrieval of information.
5. Use a pretrained hugging face sentiment analysis model to do sentiment analysis on the context of query.
6. Implementing a RAG framework to retrieve context for the query (Langchain)

Screen Shot 2022-06-03 at 11.31.35 AM.png
github URL
bottom of page