About the Project
Inspiration
Our team began with a simple—but important—insight from Sound Credit Union’s problem statement:
Many members with low or irregular income feel anxiety when interacting with their finances. They avoid checking balances, delay payments, or feel overwhelmed by uncertainty.
We realized that most fintech tools focus on optimization, while credit-union members often need something more fundamental: emotional reassurance, clarity, and supportive automation.
This inspired us to create a system that not only processes financial data, but also reduces stress, guides decisions, and automates burdensome tasks—all delivered with a compassionate tone.
That’s how our concept emerged:
A Financial Interoperability Layer (FIL) + Supportive Orchestrator that turns banking events into calm, guided member experiences.
What We Learned
Over the course of the hackathon, we explored several concepts:
1. Emotional design in fintech matters
We studied how tone, timing, and messaging can strongly influence financial decision-making. Supportive language increases trust and improves engagement (higher click-through rates in our A/B test).
2. Credit unions need orchestration—not more apps
We learned that banks already have many systems (Core, Payments, CRM, Fraud), but few tools that coordinate them. A unified data model plus an event-based orchestrator can simplify workflows dramatically.
3. Predictive modeling improves financial wellness
We implemented simple models for:
- Safe-to-Spend forecasting
- Overdraft probability
- Income vs. spending volatility
Even lightweight models can give members clearer expectations.
4. Autopay can be anxiety-reducing if paired with strong controls
Members are afraid to enable autopay due to unpredictable balances. By adding cash-flow checks and a “supportive confirmation tone,” autopay becomes a comfort feature—not a risk.
5. Privacy-first rewards are possible
We explored how optional, consent-driven “data dividends” could reward members without compromising their data.
How We Built the Project
We built a full working prototype in Streamlit, including:
✅ 1. A Common Data Model (CDM)
A normalized schema for:
- members
- accounts
- transactions
- autopay mandates
We implemented flows for:
- low balance
- anxiety balance
- lost card
- failed bill pay
- roundup savings
- smart autopay setup
3. Anxiety-Aware Intelligence
- Safe-to-Spend prediction
- Overdraft probability (logistic model) [ P(\text{overdraft}) = \frac{1}{1 + e^{-z}} ]
- Tone-optimized messaging (supportive vs. neutral)
✅ 4. Autopay + Data Dividend
Users can:
- set up monthly autopay
- preview the mandate
- choose notification preferences
- optionally opt in to a “data dividend” reward
Supportive tone consistently outperformed neutral tone in our simulation.
Challenges We Faced
1. Designing for emotions, not just analytics
Translating financial anxiety into features required understanding behavioral psychology and trust design—more nuanced than pure engineering.
2. Integrating multiple system concepts into one prototype
We had to simulate:
- a core banking engine
- payments rails
- CRM
- fraud monitoring
- a coaching/financial wellness service
All with consistent data and event flows.
3. Building a realistic, extensible orchestrator
We needed deterministic yet believable event routing, retries, and connector health states.
4. Time constraints
Balancing:
- UI
- data science
- orchestrator logic
- messaging tone
- and autopay features
within a short hackathon window was challenging but rewarding.
5. Privacy and ethics considerations
Creating a rewards mechanism without misusing personal data required careful thought. We adopted a strictly opt-in, aggregate-only model.
Conclusion
Our project demonstrates that fintech can be more than transactions—it can be emotionally supportive, predictive, and member-centric. By combining a unified data layer, event orchestration, financial wellness analytics, and empathetic communication, we built a prototype that helps members manage money with confidence and calm.
Log in or sign up for Devpost to join the conversation.