Sleeper Trades | Devpost

Inspiration

We wanted to make investing insights accessible to people who lack the time or resources to do deep research themselves. Our goal was to surface overlooked companies and give users a clear, balanced view of their potential.

What it does

Sleeper Trades collects real-time market data, news, and other public information, then uses Gemini to debate the pros and cons of investing in a given stock. The result is a concise, evidence-based analysis for the user.

How we built it

We built a sleek React frontend for a simple user experience and a Python backend to power the chatbot using Gemini’s ADK and API. We also designed a database to log requests, manage data, and support repeat queries efficiently.

Challenges we ran into

We faced long installation times for required toolkits, tricky database connection setups, and the complexity of integrating Gemini agents, all while working under hackathon time pressure for the first time.

Accomplishments that we're proud of

We delivered a full-stack product without templates or prior experience. Our agents performed strongly and quickly, providing meaningful stock insights under real-world constraints.

What we learned

We learned how to integrate multiple APIs under time pressure, troubleshoot cloud database connectivity, streamline our development workflow, and design a system that balances speed with accuracy. We also learned how to break down a complex problem into manageable parts and work collaboratively under tight deadlines.