MoneyNest

Profile screen
Chat room where you can ask questions to receive explainations along with links to interactive learning sessions to learn.
You can upload your documents which can be visualized for a better understanding
Library of pre-built and generated games, visualizations and components

Inspiration

Money stress starts early, but most personal finance tools are built for adults with stable incomes and financial vocabulary.
Teens are making real decisions now: first paychecks, spending pressure, subscriptions, credit traps, and emergency shocks.
We wanted to build something that feels supportive, not judgmental: a coach you can talk to, not a spreadsheet you avoid.
The core idea was simple: let people practice financial decisions safely before real life consequences hit.

MoneyNest is an AI-powered financial wellness platform designed to help teens build confidence through action, not lectures.
It supports multiple onboarding paths so users can start in the way that feels easiest: voice, form, document upload, or bank sync.
It turns plain chat into interactive experiences: crisis simulations, learning cards, mini-games, and personalized action plans.
It includes a teen-focused game layer (including RiskRaid) to teach risk, debt, budgeting, and resilience through consequence-based play.
It keeps progress alive across sessions with saved artifacts in a library and trackable action plans.
It adds motivation through XP and leaderboard mechanics so learning feels social and repeatable.
It adapts over time with proactive suggestions and plan refresh when a user’s profile changes.

We built MoneyNest with Next.js and TypeScript, using a modular app + API architecture.
Supabase powers authentication, profile/state persistence, and saved user artifacts.
Claude (via Vercel AI SDK stream-based orchestration) and Gemini drives generative interactions and component selection.
We use a component registry approach so AI responses can render meaningful UI blocks, not just text.
ElevenLabs powers voice and audio moments; Gemini supports generated visual learning content.
Plaid integration gives optional read-only financial context for stronger personalization.
We implemented interaction-awareness so game and simulation outcomes feed back into future AI responses.
We added staleness detection for plans so recommendations stay grounded in current user reality.

Balancing “fun” and “serious” in a finance product for teens without trivializing real hardship.
Keeping generative AI outputs consistent, safe, and useful across very different user contexts.
Maintaining continuity between deterministic pages (dashboard/plans/library) and dynamic chat experiences.
Designing game mechanics that teach real concepts instead of becoming disconnected entertainment.
Handling missing or partial user data gracefully while still delivering personalized insights.
Making a broad feature set feel like one product story instead of separate demos.

We shipped a complete end-to-end experience, not just a prototype screen or chatbot demo.
We turned AI from a “Q&A assistant” into a guided interactive system with tangible outputs and follow-through.
We created a teen-centered learning loop that combines coaching, simulation, gaming, and action planning.
We built adaptive behavior into the product: reactive answers, adaptive updates, and proactive nudges.
We made accessibility a product decision early through multi-path onboarding and human-first tone.
We built a strong hackathon narrative with real implementation depth behind it.

Teens engage more deeply with interactive decision loops than static explanations.
Personalization is only valuable when it leads to concrete next steps users can actually do.
Human tone matters as much as model quality in a sensitive domain like financial wellness.
Voice is powerful when used intentionally, not as a default for every interaction.
Product trust comes from clarity, consistency, and visible reasoning, not just “smart” outputs.
A coherent story across features is what makes innovation feel real to users and judges.

Pilot with schools, youth programs, and community organizations to validate learning outcomes.
Add longitudinal progress tracking to measure confidence growth, habit consistency, and readiness.
Expand localization and cultural personalization so advice feels truly relevant across communities.
Introduce mentor/guardian mode for guided support without removing teen agency.
Build mobile-first low-bandwidth pathways for broader accessibility.
Strengthen explainability and responsible AI safeguards around recommendations.
Deepen ecosystem partnerships (financial literacy nonprofits, insurers, youth educators) to drive adoption beyond hackathons.