SavingSpree

Inspiration

Managing finances and building better money habits is hard — especially for busy, high-achieving individuals.
We wanted to create a system that automatically helps users spend smarter, without adding stress or extra decisions to their week.
SavingSpree transforms everyday routines into opportunities for savings by predicting behaviors and integrating savings tasks directly into real user calendars.

What it does

SavingSpree connects to your financial and calendar data to:

• Predict your weekly spending patterns using machine learning
• Suggest smart actions like meal prepping, carpooling, or taking public transit
• Automatically fit those tasks into your real free time
• Export a ready-to-go weekly plan you can download or sync to your calendar SavingSpree turns free time into financial wins — helping users build consistent, smarter spending habits.

How we built it

• Backend: Python Flask API
• Finance Integration: Plaid API for sandboxed financial data
• Calendar Integration: .ics calendar parsing and generation
• Machine Learning: Random Forest model trained on synthetic user behavior data
• Scheduling Engine: Smart matching of tasks into real free slots based on user calendars
• File Handling: Export .ics files for seamless calendar imports We focused on clean modularity, allowing easy extension to future services like direct Google Calendar sync.

Challenges we ran into

• Parsing and managing real-world calendar data (.ics) with timezones correctly
• Building a machine learning model simple enough to work well with synthetic data, but extensible for real future users
• Making the task scheduling engine smart enough to prioritize preferred times (e.g., meal prep in afternoons, commuting tasks in mornings)
• Ensuring user flexibility — minimal manual input needed but maximum personalization achieved

Accomplishments that we're proud of

• Full working pipeline from finance data → behavioral prediction → dynamic calendar planning
• End-to-end automation with no human micromanagement needed
• Smart weekly plans generated with just one click
• High-quality code architecture ready for frontend and production scaling
• True combination of finance, productivity, and AI-driven personalization

What we learned

• How to integrate multiple APIs (Plaid, calendar .ics) into a single cohesive backend
• How small, smart machine learning models can drive real user impact
• The importance of scheduling logic that respects user lifestyle constraints
• That focusing on user friction removal makes apps much more powerful

What's next for SavingSpree

• Direct Google Calendar API integration (push tasks into user calendars without download)
• User authentication and multi-user support (real accounts and persistent storage)
• Expanded ML models (predict specific best tasks and expected dollar savings)
• Weekly emails or push notifications with personalized plan updates
• Frontend web and mobile app (beautiful UI for interacting with the plans) SavingSpree's goal is to make smarter spending automatic — and make building great financial habits frictionless for everyone.

Built With

• flask
• javascript
• python