Vesto

vesto

Inspiration

We wanted to create an app that can help people who struggle with basic financial knowledge, but still choose to invest based on trends on social media and word of mouth. We want to educate the next generation of investors.

What it does

Vesto helps students and first-time investors learn the fundamental analysis of investing, analyzing and breaking down each section of a 10-K report, crucial key metrics, and identifying risks before making a trade.

Learning Modules:

5 progressive modules from basic fundamentals to expert-level comparative analysis
Multiple-choice and written questions using real company data
AI-powered grading with detailed feedback across 5 criteria (100-point scale)

Stock Simulator:

Browse 20 major companies with real-time financial data
Write investment pitches and submit them to an AI Fund Manager
Build a paper trading portfolio with $10,000 starting capital
Track performance and approval rates

Progress Tracking:

Monitor completion percentage for each module
View detailed AI feedback on your answers
Track simulator performance and portfolio gains/losses

How we built it

We built the platform using Next.js 14 (React + TypeScript) with Tailwind CSS for styling, a Supabase PostgreSQL backend with real-time data and authentication, and Finnhub for live financial data. Google Gemini AI powers question generation, grading, and expert analysis, while React Hook Form and Zod handle form validation. The result is a full-stack educational fintech platform that integrates real financial data with AI-driven learning.

Technical Stack:

Frontend: Next.js 14, TypeScript, Tailwind CSS, shadcn/ui
Backend: Supabase (PostgreSQL), Next.js API Routes
AI: Google Gemini API for grading and pitch reviews
Data: Finnhub API for stock prices, fundamentals, and financial statements
Authentication: Supabase Auth

Challenges we ran into

Designing the learning flow - Figuring out how the flow of the program would work, balancing structure with flexibility, and ensuring questions build on each other while making AI feedback actionable.
AI prompt engineering - Getting consistent, structured feedback from Gemini required many iterations and careful prompt design.
Data integration - Combining Finnhub data, 10-K sections, and user progress into a cohesive experience was challenging.
Portfolio tracking - Calculating real-time gains/losses with current stock prices while maintaining performance.

Accomplishments that we're proud of

Built a fully fledged app that can help teach the next generation of investors
AI grading system that provides detailed, rubric-based feedback on student answers
Real financial data integration with live stock prices, fundamentals, and 10-K analysis
Intuitive user experience with clear navigation, progress tracking, and actionable feedback
Scalable architecture with modular code, type safety, and efficient database design

What we learned

How to leverage Artificial Intelligence to teach people effectively. This included prompt engineering for consistent AI feedback, integrating real financial data APIs, building full-stack applications with Next.js and Supabase, and designing educational experiences that balance guidance with independent learning.