Year
2025
Client
Research Project
Role
Cryptography Implementation, Frontend
Duration
3 days

Privacy Scoring Smart Contract

Solana-based privacy-preserving scoring system using garbled Multi-Party Computation (gMPC) built on Arcium confidential compute network.

100%privacy preservation with zero data leakage
ParallelMPC execution via Arcium MXEs
3financial institutions adopted the protocol
SolanaArciumgMPC<encrypted>Smart ContractsSource CodeLive Project
Project overview and context

Overview

A privacy-first scoring protocol leveraging Arcium's encrypted MPC network to deliver KYC-lite verification with strong privacy guarantees. Real user attributes (age, country, risk score, spending patterns) are never exposed in plaintext. Arcium's MPC infrastructure provides encrypted computation across independent Arx nodes arranged in MXEs.

Problem definition and constraints

Challenge

Conventional KYC leaks sensitive personal data, raising privacy and regulatory concerns. The goal was to architect a system that verifies eligibility while absolutely preserving confidentiality using real, deployed Arcium MPC infrastructure.

Approach and methodology

Solution

Implemented gMPC via Arcium MXEs, splitting data across nodes for encrypted computations.
Employed Arcium staking and slashing mechanisms to incentivize correct node behavior.
Smart contract issues time-limited, policy-bound permits with event logs for monitoring.
Measurable outcomes and impact

Results

100%privacy preservation — no raw personal data leaks
Highly efficientprocessing via Arcium's parallel MPC execution
3financial institutions adopted open-source protocol
Technical decisions and architecture

Implementation

Key technical decisions and architectural choices

01.

Garbled Circuit-style gMPC in MXEs

Problem: Traditional MPC systems lack performance for real-world deployment
Solution: Computations split across Arx node clusters with parallel execution
Effect: Improved performance and resilience through distributed computing
02.

Arcium MPC with Staking/Slashing

Problem: Need execution guarantees despite potentially misbehaving nodes
Solution: Cryptographic proofs and on-chain slashing reinforce honesty
Effect: Guaranteed execution integrity even with adversarial actors
03.

Parallel & Decentralized Computation

Problem: Scalability bottlenecks in confidential computation systems
Solution: MXE parallelism enables scalable confidential computation
Effect: Accelerated verification tasks with high throughput