30 Oct - 12 Nov 2023

As mentioned in my last update, the Aztec team has released a Request for Proposal for a decentralized prover coordination, and we've decided to submit our own proposal.

Based on all the research our team has done in the past months I have prepared an initial version of the proposal for Aztec which served as the basis.

The submission deadline was 3 November, thus during the week of 30 October our team focused on finalizing concepts, simplifying the prover design, adding additional features, preparing comparisons with other proposals, as well as addressing the questions and key considerations the Aztec team pointed out in their RFP.

In summary our decentralized prover strategy is based on

• Staking (& slashing) to create financial interest and promote honest behavior,
• Prover reputation scores to promote reliability and soft competition, and maximize liveness,
• Random prover selection from provers with top 25% reputation.

An emergency mechanism has been implemented in case of prover failure to minimize reorgs:

• N number of provers selected to generate a proof based on proof-racing,
  • The compute waste from the proof race can be considered as network redundancy to maximize liveness, and minimize the time/cost lost due to prover failure.

Other added features:

• Proof batching,
• Distributed proving,
• Integrating liquid proving pools

Our final proposal submitted to the Aztec team can be found here:
[Proposal] Decentralized Prover Network (Staking, Reputations and Proof Races)

It feels really amazing that we managed to prepare and submit our own proposal for Aztec's RFP.

This week I have been focusing on preparing slides for our team's presentation during the EPF-Day in Istanbul, as well as researching some topics more deeply and preparing replies to the questions the Aztec team asked about our proposal.