QuadraWealth

AI-powered 4-mode wealth manager: smart stock picks via RAG, sports arbitrage detection, macro-driven yield optimization, and real estate cap rate analysis — all in one dashboard.

Comment

Inspiration

As college students, we noticed that managing wealth feels fragmented — you need one app for stocks, another for savings, a spreadsheet for real estate, and yet another for spotting sports betting edges. We asked: what if one intelligent platform could handle all four? QuadraWealth was born from the idea that a unified, AI-powered dashboard could give anyone the tools to make smarter financial decisions across every asset class.

What it does

QuadraWealth is a 4-mode wealth management platform that covers the full spectrum of modern investing:

  • 📈 Stocks — A hot stock screener powered by a multi-factor scoring algorithm (P/E, ROE, EPS, D/E, Insider Holding, EMA 200) combined with RAG-based news sentiment analysis using ChromaDB. Each stock gets a composite score out of 100.
  • 🎯 The Edge — A real-time sports betting engine that scans FanDuel, DraftKings, Hard Rock Bet, and PrizePicks for arbitrage opportunities (guaranteed profit) and +EV bets. Uses Kelly Criterion for optimal bet sizing.
  • 🏦 Savings & Yields — A macro-driven yield optimizer that monitors Fed rates, CPI, unemployment, and GDP growth to dynamically allocate capital across high-yield savings, T-Bills, CDs, and I-Bonds based on your risk profile.
  • 🏠 Real Estate — A property screener that scores listings by Cap Rate, NOI, and Cash-on-Cash Return, with goal-based filtering (cash flow vs. appreciation) and interactive map visualization.

How we built it

  • Backend: FastAPI (Python) with async endpoints, background polling via SSE (Server-Sent Events), and auto-generated API docs
  • Frontend: Streamlit with a premium dark-themed UI, custom CSS animations, and interactive Plotly/Folium visualizations
  • AI/RAG: ChromaDB vector database for semantic search over financial news, powering contextual stock recommendations
  • Data: yfinance for live stock data, The Odds API for real-time sports odds, and Realty Mole Property API for real estate data
  • Deployment: Railway for production hosting with separate frontend/backend services

Challenges we ran into

  • Real-time polling at scale: Keeping odds data fresh across 6 sports and 4 sportsbooks without hitting API rate limits required building a custom async poller with SSE streaming
  • RAG accuracy: Tuning ChromaDB embeddings so stock news sentiment actually correlated with better picks took several iterations
  • Arbitrage math: Handling edge cases where implied probability calculations break down with extreme odds or when sportsbooks update lines mid-scan
  • Deployment: Orchestrating a dual-service architecture (FastAPI + Streamlit) on Railway with proper environment variable management

Accomplishments that we're proud of

  • Built a fully functional arbitrage scanner that detects guaranteed-profit opportunities in real-time
  • Implemented a production-grade RAG pipeline for AI-powered stock recommendations
  • Created a premium, Apple-inspired UI with smooth animations and glassmorphism effects
  • Deployed a live, working application accessible from anywhere

What we learned

  • How to build and deploy RAG (Retrieval-Augmented Generation) systems for financial analysis
  • The mathematics behind sports betting arbitrage and the Kelly Criterion
  • Real estate investment analysis: Cap Rate, NOI, Cash-on-Cash calculations
  • Macro-driven asset allocation — how Fed policy, inflation, and yield curves affect savings strategies
  • Production deployment patterns with FastAPI + Streamlit on Railway

What's next for QuadraWealth

  • Live portfolio tracking — Connect brokerage accounts via Plaid for real P&L
  • ML price prediction — LSTM models for stock and real estate price forecasting
  • Push alerts — Notify users instantly when an arbitrage window opens
  • Social features — Share and compare portfolios with other users

Built With

Share this project:

Updates