PRIVØRA
Private Capital for Public Impact
Created on 4th December 2025
•
PRIVØRA
Private Capital for Public Impact
The problem PRIVØRA solves
1. The "Transparency Trap" in Philanthropy
We face a dangerous paradox: builders and activists in high-risk jurisdictions need funding the most, but they cannot afford to be seen. Current tools force a deadly choice: Doxx yourself (GoFundMe/KYC) or leak your financial graph (Public Blockchains).
Privora's Solution: We decoupled Verification from Identity. We use Reclaim Protocol to generate Zero-Knowledge proofs of real-world attributes (like GitHub commits or Location) without revealing the user's identity. Combined with NEAR AI (TEEs) to sanitize bio data and Zcash for shielded payments, we allow capital to flow to verified builders without exposing the humans behind the code.
2. The "Public Reputation" Dilemma (FHE)
In Web3, reputation is usually public. If you receive a "Scholarship SBT" or a "Patron Badge," the whole world sees it. For privacy-centric communities like the Network School, a public list of residents is a security vulnerability.
Privora's Solution: We built Encrypted On-Chain Reputation using Fhenix.
Our ImpactSBT contract mints Soulbound Tokens where the donation amount is encrypted (euint128). Patrons can prove they supported the cause without revealing how much, preserving the privacy of their wealth.
Our NetworkSchoolVerifier stores residency status as an encrypted boolean (ebool). The contract knows who is a member, but the public blockchain explorer shows only encrypted bytes. This allows for a "Dark Census"—we can verify and count the population without doxxing the list of members.
3. "Vetting Fatigue" for Donors
Donors want to help but lack the time to manually verify hundreds of anonymous profiles to distinguish real builders from scammers.
Privora's Solution: We introduced Agentic Philanthropy. Instead of manual browsing, Donors use natural language to command an AI Agent ("Fund verified anti-censorship devs"). The Agent uses TEE-based inference to semantically match the Donor's intent with the Reclaim-verified anonymous personas, ensuring funds only go to candidates who have mathematically proven their skills and location.
Architecture Diagram
Challenges we ran into
1. The "Reputation Leaks Privacy" Paradox
We wanted to gamify philanthropy with Soulbound Tokens (SBTs), but realizing that a standard ERC-721 token would doxx the shielded Zcash transaction was a major hurdle. If an address holds a "1000 ZEC Donor" badge, the privacy of the Zcash layer is broken by the Ethereum layer.
How we solved it: We moved the logic to Fhenix. We wrote the ImpactSBT.sol contract using Fully Homomorphic Encryption. Instead of storing uint256 amount, we store euint128 amountEnc. This allows the contract to track total impact mathematically without ever revealing the plain-text values to the public network. Debugging the type conversions between Solidity uint and FHE euint was a significant learning curve.
2. Bridging Stateless Inference with Stateful Money
The NEAR AI TEE is a stateless inference engine—it processes text and outputs JSON, but it cannot "hold" a wallet or sign transactions. We needed a way to turn the AI's matching advice into a secure financial action.
How we solved it: We implemented a QR-Based Handshake. The TEE performs the matching logic and outputs a JSON decision. Our frontend immediately converts this decision into a Payment QR Code (complete with the recipient's shielded address and an encrypted memo). This allows the AI to drive the logic while keeping the cryptographic signing keys air-gapped on the user's mobile Zashi wallet.
3. Verifying "Invisible" Residents
We needed to verify Network School members on-chain without creating a public "hit list" of addresses. A standard whitelist contract (mapping(address => bool)) is publicly readable, which defeats the purpose of a privacy society.
How we solved it: We implemented NetworkSchoolVerifier.sol on Fhenix using mapping(address => ebool). By storing the verification status as an encrypted boolean, we created a registry that is programmable but opaque. The contract logic can check membership, but an external observer cannot scrape the list of members. Combining this FHE logic with the off-chain Reclaim Protocol proofs required careful orchestration to ensure the "Verified" status was updated securely.
Tracks Applied (3)
General Bounty
Network School
Private DeFi & Trading
Zcash Community Grants
Privacy-Preserving AI & Computation
NEAR Protocol
Cheer Project
Cheering for a project means supporting a project you like with as little as 0.0025 ETH. Right now, you can Cheer using ETH on Arbitrum, Optimism and Base.