Inspiration
Financial and macroeconomic indicators are increasingly dependent on geopolitical factors due to increasing multipolarity and conflicts. This calls for a universal state machine allowing to formalise the impact of a news to defined economic indicator like oil prices.
What it does
Input any question or market and the engine will source and weight the most relevant signals for that question.
How we built it
We used a Hawkes framework couple with a multi-agent system to set weights on signal and aggregate them into
Challenges we ran into
Addressing the heterogeneity of data sources in a common framework. Building an event driven state machine for the world cup is very different from building one for geopolitical events.
Accomplishments that we're proud of
We managed to pool both open source twitter data and gdelt data and construct a coherent weight setting framework managing the heterogeneity of geopolitical markets.
What we learned
We need to learn to calibrate the weight to given market dynamics.
What's next for Eversight Signal
We will release this api to prediction market traders next week.
Log in or sign up for Devpost to join the conversation.