Inspiration
we saw people using polymarket to bet on things like elections and fed decisions. we realized this data shows what people believe about the future. so we thought, what if an ai could read those beliefs and find which stocks might move next.
What it does
our ai agent reads live polymarket trends. it finds what people talk about most, like rate cuts or inflation. then it connects those trends to stock data from databento. if the crowd expects a rate cut, it looks for stocks that went up when rates dropped before. it gives the top three stocks in one sector or a simple yes or no for one stock. everything happens in one chatbot.
How we built it
we built it in python using openai, polymarket, and databento. one part reads your question, another finds trends, another links them to stocks, another checks how stocks reacted before, and one gives the final answer. we used streamlit for the chat demo.
Challenges we ran into
polymarket had no direct data for healthcare, so the ai had to find related topics like policy or inflation. databento gave too much data, so we had to make it run faster. the hardest part was making the ai link crowd belief to stock moves.
Accomplishments that we're proud of
we finished a working demo in a weekend. the ai gives real reasons, not random picks. we connected two worlds, prediction markets and stocks, in a clear system. the demo feels simple and fun.
What we learned
we learned that crowd belief data can be useful if you connect it to real market data. we also learned that ai can reason well with just clear logic and apis. keeping it simple made it work better.
What's next
next we want to add more sectors and smarter trend tracking. we also plan to let users ask questions like what happens if rates go up and see a real answer from both polymarket and databento data.
Built With
- polymarket
- python
- react
- typescript