Details

InvestSight is a conceptual web platform made for the 2020 PolyFinTech100 API Hackathon that aims to generate succinct analyses of earnings call transcripts from corporations on the US stock market.

This project was helmed by a group of 6 people, also named Team AGILE:

• Krystal (Team Lead / Presenter)
• Ethan (UX/UI + Presentation Designer)
• Dylan (Research / Concept Design)
• Yun Lin (UX/UI Concept Designer)
• Evan (Data analyst / Research)
• Jordan Khong (me) (Programming / Research)

The project was planned to have this workflow:

• Gather Raw Data from Earning Call Transcripts
• Clean Data
• Use of TF-IDF (term frequency–inverse document frequency) to extract word importance
• Use a MonkeyLearn API to determine Sentiment
• Calculation of Sentiment using Custom Formula
• Show a dashboard of good / bad sentiment breakdown and comparisions to previous quarter scores from the same company

You may view more implementation details in the Proposal Presentation