AI Agent for Financial Trend Forecasting with bunq API

AI Agent for Financial Trend Forecasting

Project Overview

The project consists in an intelligent agent built on top of the bunq API that transforms raw transaction data into actionable, year-ahead financial insights:

Trend Prediction: Using customer transactions, we trained a logistic regression model to forecast personalized rent, salary, and subscription patterns.
Contextual Validation: Our agent performs web searches to compare in-house trends against global market trajectories, ranking and delivering informed expense projections.
Data-Driven Discoveries: We uncovered how deeply insightful banking data can be—especially when enriched with geolocation data, sparking conversations on personalization and privacy.

LinkedIn Post

To read the full story on LinkedIn:

Demo Video

Watch the live demo presentation from the hackathon:

Try the Demo Yourself

🎯 Launch the live demo here:
👉 https://riccardocampanella.github.io/bunq_hackathon/

Key Takeaways

Rapid Prototyping: From concept to working product in 24 hours.
Hybrid Intelligence: Combined symbolic ML (logistic regression) with contextual web intelligence.
User Engagement: Novel UI/UX through humor and memes.
Ethical Reflection: Leveraging sensitive data responsibly while enhancing personalization.

Riccardo Campanella