NOVA | Devpost

Project Story Inspiration Scotiabank already owns the relationship with 6 million young Canadians aged 18–34. Yet, over the last decade, Wealthsimple has built a $100B AUM empire largely by winning over this exact demographic.

The problem isn't the product, Scotiabank offers excellent investment vehicles. The problem is timing and perception . During tax season, fintech competitors deploy broad, aggressive push notifications reminding users to invest their refunds. By pure coincidence, these marketing blasts hit right around the time a user's CRA refund lands, capturing their attention through lucky timing.

Scotiabank doesn't need to rely on coincidence. We see the exact second that deposit hits the user's checking account, yet traditional systems say nothing. We built NOVA to turn what competitors do via lucky timing into a precise, real-time data weapon.

What it does NOVA is an investment activation layer built directly into the existing Scotiabank mobile app. Instead of forcing users into a complex wealth management ecosystem, it meets them at the exact moment of peak financial intent through three core components:

The Moment Engine: While other apps blast generic, timed marketing campaigns hoping for a coincidence, NOVA monitors data streams in real-time to detect 12 specific financial life events (eg, tax refunds, salary increases, clearing a debt). The exact second year event triggers, it fires a highly contextual, immediate notification based on absolute certainty, not a guess.

The 47-Second Flow: Eliminates onboarding friction. Because Scotiabank already possesses the user's legal name, SIN, income data, and KYC information, we bypass the traditional onboarding nightmare. Users answer just two simple questions to open a TFSA.

AutoGrow: Shifts the psychological friction of investing by defaulting new accounts to a $25/month recurring contribution. By switching from an opt-in model to an opt-out model, we project massive, stable asset growth.

How we built it We built NOVA as a modular addition designed to sit seamlessly on top of legacy banking frameworks without requiring an infrastructure overhaul.

Backend & Analytics: We simulated a real-time banking ledger using Python and FastAPI. The Moment Engine utilizes a pattern-matching pipeline that monitors incoming transaction descriptions (like FED-PROV CRA-ARC) to trigger instant webhooks.

Frontend Experience: The mobile UI was built using React Native (simulated via Vite/React for our frontend demo) to mimic the native Scotiabank app environment, focusing heavily on micro-interactions during the 47-second onboarding flow.

Data Layer: We utilized a lightweight database schema to map existing user profile data (SIN, KYC, address) directly into the new investment account creation payload, proving that zero-friction account opening is entirely possible with data the bank already holds.

Challenges we faced Moving from Broad Marketing to Real-Time Triggers: The biggest technical hurdle was building an engine that could parse transactional data streams instantly without creating performance bottlenecks. We had to ensure our event-listener logic could distinguish a standard deposit from a landmark financial life event in milliseconds.

Designing the 2-Question Compliance Friction: We had to balance absolute simplicity with regulatory compliance. Determining which legal hurdles could be safely automated using pre-existing bank data versus what required explicit user confirmation was a delicate product design challenge.

The Psychology of Defaults: Engineering the AutoGrow flow requires careful UI/UX positioning. We wanted to ensure the $25/month default felt like a helpful, protective financial guardrail rather than an aggressive upsell.

What we learned We learned that the ultimate competitive advantage for traditional institutions isn't a prettier UI, it is the permanent data moat . A fintech app has to rely on generic seasonal marketing blasts, hoping it aligns with a user's pay cycle or tax return. Scotiabank sees the wealth created at the exact second it happens.

By replacing "coincidence marketing" with absolute data certainty, and eliminating redundant data entry, we can capture billions in leaked AUM.