Financial news research assistant

The brief

Equity analysts read dozens of news items a day about the companies they cover. Build an assistant that collects news, classifies sentiment with a finance-tuned model, summarises what changed, and links it to price moves. Good looks like: sentiment accuracy on financial text, useful daily summaries, and an analyst-facing app.

What you will build

1. Fine-tune or prompt a model for financial sentiment and evaluate it
2. Group news by company and theme
3. Generate daily summaries and check them for errors
4. Link news to price moves
5. Build the analyst app
6. Report accuracy, risks of errors and how to use it safely

Free, open data to build it with

• Financial PhraseBank ↗ Hugging Face
• NIFTY-50 Stock Market Data ↗ Kaggle

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.