FirstAdvisor

FirstAdvisor turns financial documents into simple, personalized insights to help young adults make smarter money decisions.

Comment 1

Inspiration

We built FirstAdvisor because many young adults enter the financial world without being taught how it works. Things like taxes, benefits, leases, credit cards, retirement plans, and student loans are often explained using confusing language that assumes prior knowledge.

As students in similar situations, we saw how financial literacy gaps can create long-term stress and expensive mistakes, especially for lower-income communities and people entering their first jobs. Many people do not know what questions to ask, which makes financial systems feel intimidating and inaccessible.

We wanted to create a tool that explains financial information in simple language before small mistakes become major financial problems.

Our goal was to build something that feels supportive, understandable, and personalized - especially for people navigating these systems for the first time.

What it does

FirstAdvisor is an AI-powered financial literacy assistant for young adults.

Users first complete a short onboarding form with information such as:

  • income
  • employment type
  • state
  • financial goals
  • student loans
  • benefits information

They can then upload documents like:

  • pay stubs
  • credit card agreements
  • leases
  • loan offers
  • job offer letters

The AI analyzes the document and explains important sections in simple language. It also looks for potential risks such as:

  • high APRs
  • overdraft risks
  • expensive leases
  • missing employer retirement matches
  • predatory terms
  • suspicious clauses

The main feature is our Financial Blind Spot Detector.

Instead of analyzing documents generically, the app combines the uploaded document with the user’s onboarding profile to generate personalized insights.

For example:

  • warning a user that their rent may be too high for their income
  • identifying that they are missing out on employer 401(k) match contributions
  • explaining why a high-interest credit card may be especially risky given their existing student loans

We also included:

  • future financial impact visualizers
  • financial term glossary pages
  • personalized dashboards
  • sample documents for practice and learning

How we built it

We built the frontend using HTML, CSS, and JavaScript.

We used Firebase for authentication, user sign-in, and storing onboarding and session data.

We used the Gemini API to generate personalized explanations and financial insights from uploaded documents.

The app combines onboarding profile information, extracted document content, rule-based logic, and AI-generated explanations to create personalized financial recommendations.

We also designed a custom dashboard interface with:

  • document history
  • financial statistics
  • blind spot analysis
  • future impact tools
  • profile summaries

Claude helped generate parts of the website UI and layout design. ChatGPT helped with brainstorming, testing ideas, and logo generation.

Challenges we ran into

One challenge was setting up and managing API keys securely while working across multiple services and environments.

We also spent a lot of time implementing the technical pipeline between file uploads, document parsing, onboarding data, and AI-generated responses.

Authentication and user session management with Firebase also took time to fully connect across pages.

Another challenge was balancing personalization with ethical concerns. We wanted the AI to provide meaningful guidance without making users feel judged or pressured.

We also had to think carefully about financial bias, explainability, transparency, and simplifying complex financial language while still keeping the recommendations useful.

Accomplishments that we're proud of

We are proud that we created something with real social impact that could genuinely help our peers and communities.

Financial literacy tools are often overwhelming, generic, or designed for people who already understand financial systems. We built something that explains information in a way we personally would have wanted when entering adulthood.

We are especially proud of the Financial Blind Spot Detector because it makes the experience feel personalized rather than generic. The app understands the person behind the document instead of only analyzing the document itself.

We are also proud that we centered ethical considerations such as fairness, transparency, accessibility, and financial privilege bias throughout the design process.

And of course, we are extremely proud of the bear mascot!!

What we learned

We learned a lot about financial systems and terminology while building this project, including retirement plans, employer matches, APRs, loan structures, taxes, and much more.

We also researched the broader impact of financial literacy inequality and how lack of access to financial education disproportionately harms lower-income and first-generation communities.

On the technical side, we learned more about Firebase authentication, API integration, and document-based AI workflows since it was our first time working with some of the tools.

We also learned how important ethical considerations are when working with AI and financial information, especially regarding transparency, fairness, human oversight, and data privacy.

What's next for FirstAdvisor

With more time, we want to expand FirstAdvisor into a long-term personalized financial guidance platform rather than only a document analyzer.

One major feature we want to improve is the Financial Blind Spot Timeline.

Instead of isolated warnings, users would be able to track:

  • financial improvements over time
  • repeated risky behaviors
  • long-term consequences
  • progress toward goals

For example:

"You have only made minimum credit card payments for 3 months."

We also want to improve transparency and explainability by adding:

  • “Why am I seeing this?” panels
  • confidence levels
  • AI reasoning summaries
  • direct links between uploaded documents and recommendations

This would help users better understand how the system generated each insight and prevent the AI from feeling like a black box.

Another major goal is improving accessibility and inclusivity through:

  • multilingual support
  • simplified reading modes
  • plain-language financial rewrites
  • interactive glossary explanations inside uploaded documents

We want FirstAdvisor to help reduce vocabulary gatekeeping and make financial systems easier to navigate for everyone, regardless of background or prior financial knowledge.

Built With

Share this project:

Updates