Skip to main content

Technical Concepts

Before diving into the implementation, it helps to understand what an OPRF is and how TACEO:OPRF extends it.

What is a Pseudo-Random Function (PRF)?​

A Pseudo-Random Function (PRF) is a deterministic function that produces outputs that look random to anyone who doesn't know the secret key.

In practice, PRFs are implemented using:

  • Block ciphers (like ChaCha or AES)
  • Hash-based constructions (such as HMAC)

Given a secret key k and an input x, a PRF deterministically computes:

F(k, x) → random-looking value

What is an Oblivious PRF (OPRF)?​

An Oblivious PRF (OPRF) is a special kind of PRF computed jointly by two parties:

  • Client: Provides the input x
  • Server: Holds the secret key k and evaluates the function

The key property is mutual privacy:

  • Client never learns the key k
  • Server never learns the input x

How TACEO:OPRF Extends This​

TACEO:OPRF builds on the OPRF design with threshold security:

  • The secret key k is secret-shared across multiple parties
  • No single party can evaluate the PRF alone
  • A threshold of parties must cooperate
  • Provides MPC trust guarantees while keeping inputs and keys private

Architecture Overview​

The TACEO:OPRF architecture consists of:

  1. Independent OPRF nodes that hold shares of the OPRF secret key(s)
  2. On-chain public registry of corresponding OPRF public keys
  3. Authorized accounts that can trigger secure key generation

TACEO_OPRF.png

Standard Protocol Flow​

Here's how a user interacts with TACEO:OPRF:

1. Query Generation​

The user generates a blinded OPRF query and produces an authorization proof.

2. Node Processing​

OPRF nodes:

  • Receive the query
  • Verify the authorization proof
  • Contribute with their key-share

3. Response Aggregation​

The client:

  • Receives responses from threshold-many nodes
  • Combines answers into the full OPRF output

4. Optional Proof Generation​

The client can prove valid OPRF execution in a ZK proof and submit both the output and proof to a verifier.

Key Security Properties​

Privacy​

  • Input privacy: Your secrets are never revealed
  • Key privacy: No single node knows the complete key
  • Computational privacy: Intermediate steps remain hidden

Integrity​

  • Threshold security: Requires cooperation of multiple honest nodes
  • Verifiable computation: Results can be proven without revealing inputs
  • Authorization enforcement: Custom eligibility requirements

Flexibility​

  • Custom authorization: Implement project-specific access control
  • ZK-friendly design: Compatible with zero-knowledge proof systems
  • Multi-use keys: Support multiple OPRF instances per authorization module

Further Reading​

For deeper technical details: