Dike

This project introduces DIKE Protocol — The Multiverse Market, a next-generation prediction market that reimagines how capital flows across multiple bets. Traditional prediction markets force users to split their capital between predictions, limiting upside. DIKE changes that by introducing prediction trees, where a user can start with one stake and then leverage up to 60% of their initial capital to create “child” predictions without needing additional funds. This creates a compounding effect where wins branch into further wins, multiplying outcomes far beyond the usual 2x ceiling.

To power this system, DIKE uses an AMM-based market-making mechanism, allowing users to enter and exit prediction positions at fair, liquid prices without depending on order books. Liquidity providers back the pools, and user trades dynamically adjust odds, ensuring continuous price discovery. On top of this, we embed a human validation layer: prediction outcomes are resolved not only by smart contract logic but also by KYCed human validators verified using Self (a decentralized identity/KYC solution). This ensures that resolutions are trustworthy, tamper-resistant, and free from Sybil attacks, while still being decentralized.

The leverage provided for branching is extended by the protocol itself as a loan. Once a prediction resolves, the loan plus interest is reclaimed, keeping the system self-sustaining and capital-efficient. If a user’s overall position falls below the loan + interest threshold, liquidation occurs to protect the protocol. This balance of risk-managed leverage, AMM-driven liquidity, and human-verified outcome resolution creates a uniquely resilient ecosystem.

By combining efficient capital use, decentralized liquidity, and trusted resolution mechanisms, DIKE transforms prediction markets from static wagers into living multiverse strategies, where users can explore multiple outcomes, amplify their returns, and rely on a fair, transparent settlement process.

We built DIKE Protocol by combining on-chain contracts, decentralized price feeds, stable payment rails, and identity verification into one seamless system.

At the core, our smart contracts (Solidity, deployed on Sepolia) handle prediction markets, branching logic, and loan accounting. We integrated Pyth Network price feeds to resolve price-action–based predictions without relying on human oracles, ensuring accurate, real-time, and tamper-proof data. For non-price outcomes, we use human validators, but with a twist: they are KYCed using the Self SDK. Validators scan a QR code through Self, which routes to our backend verifier, ensuring they are unique, verified individuals. This reduces Sybil risks and guarantees trust in human-resolved predictions.

To keep the trading environment stable, we chose PayPal’s PyUSD stablecoin as our base currency. By anchoring trades in PyUSD, users don’t face volatility risks if the protocol’s native token dips, creating a more resilient and trustworthy market.

On the front end, we built a responsive Next.js application, which connects seamlessly to our contracts and validator flows. The backend ties together Self’s verification process and feeds validator results into the smart contracts. For the AMM engine that powers predictions, we built custom logic that dynamically adjusts odds and liquidity without relying on centralized order books.

A notable hacky piece was getting Self’s QR-based KYC approval flow to sync with our contract-level validator registry — bridging off-chain verified identities with on-chain validation in near real-time. All this while trying to get the collaterals right and getting the predictions to branch out was pretty daunting but nonetheless, exciting!

Together, these technologies create a robust, capital-efficient prediction market with multiverse branching, verifiable outcomes, and stable trading mechanics.