ZKombinator
Verifiable Multi-Party Computation to secure DeSci research funding
Demo
About
The problem it solves Currently, DeSci research involves multiple institutions that collaborate to perform research. Over time there have been several ways to perform DeSci with methods like Homomorphic Encryption and Federated ML being popular examples of performing DeSci research. However, amongst other problems, seeking funding is still hard, mainly because the cluster of the institutions working together cannot reveal their sensitive research information (revenue, breakthrough, methodologies), either publicly or to funding institutions. As a cluster, they have enough potential for funding, but splitting the revenue across the institutions to seek funding without revealing their internal sensitive information is a challenge. To solve this, we introduce ZKombinator, where we perform multiparty computation across various organizations, research labs, instiutions working together in clusters for DeSci research. We use **Zero knowledge proofs** to verify their claims and use BGW (Ben-Or, Goldwasser and Widgerson) algorithm to calculate the fair funding split among the cluster without revealing their sensitive documents. While zk proofs allow institutions to verify their claims, its still extremely hard to perform calculations on internal information. Methods like homomorphic encryption are not suitable in this case because of their linear nature. To overcome this, we use the BGW algorithm to securely calculate the funding to be received by each startup without transfer of any sensitive information across any device in a decentralized manner. We ensure that none of the nodes are adverserial **(that is, they haven't tampered with the arithmetic circuit calculation)** by generating zero knowledge proof of calculation done by each node (institution) and verify it via zKVerify. This ensures that all of the parties have followed the algorithm, without tampering any data, and the parties securely receive funding without transfer of any senstive information among any parties.
How it's made
Challenges we ran into This project was very new to us. As a team, none of us had ever dealt with Secure Multi-Party Computations. Hence, it was a unique challenge and getting started was a good challenge, which helped us implement the BGW algorithm with Zero Knowledge as the core of our project. We also faced hurdles with intergrating the *Arbitrum* smart contract.