AADIL

• 

  dashboard

• 

  planning

• 

  budgeting

• 

  payments

• 

  reports

Inspiration

College students think in semesters, not isolated days. The real questions are “Will I be broke by finals?” and “Can I afford Spring Break and still eat?” Traditional banking and budgeting apps show a current balance and past transactions, but they do not explain the semester-long impact of everyday choices like coffee, takeout, or rideshares. Students often realize there is a problem only when a term bill is due or their account is already low.

On top of that, campus life has its own patterns that generic finance tools ignore. Students juggle meal plans and off campus food, irregular income from part time jobs or refunds, and large, infrequent charges such as tuition, textbooks, and fees. These do not align neatly with standard monthly budgets. AADIL is motivated by this gap: students need a tool that understands the rhythm of a semester and translates daily spending into a clear picture of where they will stand when it matters most.

What it does

AADIL is an AI powered smart wallet built around the structure of a semester. Students connect their financial data, and the system learns when and how they spend, distinguishing between flexible spending (off campus food, coffee, entertainment, rideshare, groceries), fixed obligations (phone bill, utilities), and academic costs (textbooks, tuition, supplies).

Using that understanding, AADIL shows both a real time balance and a projected balance at the end of the semester based on current habits. Students can simulate a day to see how repeated choices affect their future, set goals in a dedicated plan tab, track weekly budgets by category, review upcoming charges and subscriptions in a payments style view, and read AI generated reports that summarize recent behavior and suggest clear improvements.

How we built it

We structured AADIL around a backend service that models student finances as a timeline across the semester. All transactions are stored with core attributes such as date, amount, category, and type (income, expense, meal swipe, term bill), and the system walks forward day by day from an initial balance to compute where the student stands at any point in time.

To project into the future, AADIL looks back at history and learns simple patterns, such as which days tend to see more spending, which categories dominate, and when recurring income arrives. From this it generates two projected paths: one that extends current behavior and one “AADIL plan” path that assumes more intentional choices in flexible categories while keeping fixed and academic costs unchanged. On top of this, we added an AI layer that receives summarized spending data and returns concise insights, weekly recaps, and a semester overview in plain English.

Challenges we ran into

A key challenge was designing projections that feel realistic yet easy to understand. Projections that were too rigid looked artificial, while projections that were too random did not feel trustworthy. We had to strike a balance between reflecting historical patterns and adding just enough variability, and present the result in a way that a non technical user could interpret at a glance.

Another challenge was using AI in a way that adds clarity rather than noise. Long, generic advice is easy to generate but rarely useful, so we iterated on prompts and formatting until the insights were short, specific, and respectful of how sensitive money can be. In the hackathon context, we also needed to avoid slow, repeated AI calls, so we shifted generation to a separate step that runs up front and stores results for the main application to serve quickly.

Accomplishments that we're proud of

We are proud that AADIL operates as a complete, usable system rather than a collection of ideas. A student can start with an initial balance, add or import transactions, advance the simulated semester day by day, and watch their projected balance respond. Seeing the current path and the AADIL plan side by side makes the impact of modest spending changes tangible.

We are also proud of how cohesive the experience feels. The dashboard, timeline, goals, budgets, payments view, and AI reports reinforce one another, so students see current status, future trajectory, and actionable advice in one place. The product is grounded in real student life, with categories and examples that match actual behaviors like dining choices, rideshares, streaming subscriptions, and term bill deadlines, and early feedback has been that it is something students could realistically imagine using.

What we learned

Building AADIL taught us how to turn a shared, somewhat vague problem into a concrete product. We learned to frame student finances in terms of the semester cycle, not just monthly budgets, and to combine familiar concepts like transaction histories and timelines with an AI assistant that adds context without overwhelming the user.

We also learned to think more like product designers than just engineers. Tone, clarity, and flow mattered as much as implementation details. Crafting projections that are easy to read, insights that are encouraging rather than judgmental, and screens that guide users from information to action showed us how thoughtful financial feedback can reduce stress and help students feel more in control.

What's next for AADIL

The next step for AADIL is to connect it to live financial data. We plan to integrate a secure account linking solution so students can sync real bank and card transactions, and explore connections to campus systems such as meal plan balances, term bill portals, and student card transactions. This will allow AADIL to represent a complete, real time view of a student’s financial life. We also intend to build a polished mobile experience so checking projections and weekly reports fits naturally into a student’s routine. Over time, we want AADIL to adapt to individual patterns and priorities, and eventually expand beyond Rutgers into a student focused financial wellness platform that campuses can deploy to support their communities at scale.

Built With

• expo.io
• fastapi
• numpy
• openaiapi
• python
• react-native
• uvicorn