The idea came from something i kept running into while playing with trading bots. Once your bot needs access to more than 1 venue, you suddenly have to deal with different market schemas, contract definitions, prices, fees and APIs before you can reason about where to trade. That gets messy real fast, especially with prediction markets. The same real-world event can exist on Kalshi & Polymarket, but the contracts can be represented differently.
Here's my broader thesis: as more trading becomes automated, bots/agents will need a common interface to access different financial markets. Prediction markets felt like a good place to start because the fragmentation is obvious.
With Layer's SDK, you can: - pull normalized prediction-market data - match equivalent markets across Kalshi and Polymarket - compare prices across venues - see where a trade has a better payout - identify potential cross-venue arbitrage opportunities
Additionally, I built an example bot called Spread using Layer. It finds matched markets across venues and compares their prices after fees, so you can see which venue pays more for the same bet.
Layer is still super early and i would love your feedback, especially from people building trading bots/agents, or prediction-market tools.
Feel free to break it, build on it. Would love to see what you do with it.