Inspiration
63% of Americans live paycheck-to-paycheck, yet financial literacy is rarely taught in schools. When beginners open trading apps like Robinhood, they risk real money in a system they don't understand. We asked: How do people learn markets without fear, pressure, or financial loss?
Technical Inspiration: We saw what was possible with modern AI (e.g., StockBot's real-time market analysis powered by Llama 3.3-70B, and TradingView widgets) and realized the same tech could serve education, not just trading.
Educational Philosophy: Inspired by Socrates, history's greatest teacher, we realized the answer wasn't more predictions but rather better questions. Stocrates takes the technical power of AI + live market data and redirects it entirely toward teaching critical thinking, not making trading decisions.
The Key Difference:
- StockBot & similar apps: "Here's a live stock chart and analysis" (useful for traders)
- Stocrates: "Why did this happen? What patterns do you see? What would you do?" (teaches investors to think)
We combined Socratic questioning with real historical market data to teach users how to analyze markets, not speculate on them.
What We Built
Stocrates is an AI-powered financial literacy platform that teaches market behaviour through:
- Event Analysis - Asks questions like "Why did Tesla drop?" and shows real news, community sentiment, and historical market reactions
- Pattern Recognition - Teaches 5 major chart patterns (Breakout, Fakeout, Retest, Head & Shoulders, Continuation) with historical success rates (55-72%)
- Paper Trading - Time-travel to the past and practice with fake money using real historical prices
- Socratic AI - Instead of predictions, the AI teaches
How we built it
We built Stocrates using a modern web stack focused on speed, transparency, and education.
AI: Groq’s Llama 3.3-70B (pre-trained, open-source)
- Used only for inference (no custom training)
- Guided by strict educational prompts using the Socratic method
- Prevents predictions and financial advice by design
Data Architecture:
Integrated 5 data sources with a 7-layer fallback system for 99.9% uptime
Finnhub → NewsAPI → Reddit → Cache ensures service continues even when APIs rate-limit
Polygon.io provides intraday candles for precise pattern detection
Pattern Analysis:
- Deterministic, rule-based engine (not machine learning)
- Compares historical price movements around similar events
- Shows historical success rates as context, not forecasts
Challenges we ran into
API rate limits & reliability
Financial data APIs are heavily rate-limitedBalancing clarity with complexity
Markets are complex, but the experience needs to remain approachable for beginners.Ethical boundaries
We intentionally avoided features common in trading apps to keep the focus on learning, not speculation.Technical Issues and Learning Curve For part of the team, the frameworks and toolings were new, so understanding how everything connected was a challenge, especially under a hackathon timeline.
Accomplishments that we're proud of
- Successfully combined ancient teaching methods with modern AI
- Created a risk-free paper trading experience using real historical prices
- Designed a system that is transparent and honest about its limitations
- Built a fully working educational platform in under 48 hours
What we learned
- How important accessible financial education really is
- Deeper understanding of how markets react to events
- AI is most powerful as a teacher and guide, not an oracle
- Ethical design in fintech requires clear, enforced boundaries
- Hands-on experience with APIs, modern frameworks, and system architecture
- The importance of researching what is ethical (and what should be avoided) in financial technology
What's next for Stocrates
- More historical scenarios and global market events
- Classroom tools and educator dashboards
- Personalized learning paths based on user curiosity (e.g., halal stocks)
- Accessibility support: Different languages, text-to-speech, etc.
“Learn markets through patterns, not predictions.”
Built With
- css
- finnhub-api
- finnhubapi
- groq-api-(llama-3.3-70b)
- json
- newsapi
- next.js
- polygonapi
- react
- reddit-data
- stock-market-data
- tailwind-css
- typescript

Log in or sign up for Devpost to join the conversation.