Current State of Team

In Winter 2021, we built scrapers and sentiment models for Twitter, Reddit, Wall Street Journal, and quarterly investor calls. It's not fully working, so one goal of this semester is fixing it.

We'll focus on delivering real value to the Platform and Investing Side. This means scraping posts from different social media platforms and using pre-built sentiment models to analyze them (two more goals: building scrapers and finding models).

We'll wrap up our code in an API, which will be deployed on Google Cloud Functions to work with the Platform's stack. We want to store sentiment over time. This means picking a basket of stocks and scraping and analyzing posts about them on multiple platforms each day. These will be stored in Firestore, a NoSQL Google Cloud database.

In summary, here are the goals:

1. Fix W2021 algorithm (Reddit, WSJ, Twitter, quarterly calls)

2. Choose new social media platforms and build scrapers

3. Find and use existing sentiment models

4. Deploy scrapers and models as API

5. Call APIs every day on a basket of stocks and store sentiment for those stocks