We work on cryptography and protocol design heavy projects.
We have three lines of work: Protocols, Research, Products.
Storage Provider Costs
Reducing Proving costs
Reducing Deal Making costs
Storage Market Programmability
Make Filecoin easy to program with better standards and programmable storage markets features
Export Filecoin storage to other Web3 systems, making it useful from anywhere.
Data Preparers Markets
A large majority of users onboard data onto the Filecoin network via an aggregator. This FRC makes aggregators’ work verifiable.
Economic Stability & Robustness
Pledge Collateral Scarcity
Reduce pledge collateral requirements when tokens are hard to lease.
Network value capture
Align Filecoin network fees with value provided to participants.
Monitor the network, detect issues and propose immediate solutions.
Lookup Arguments for SHA
Filecoin Storage Proofs circuits are for more than 90% SHA. By making SHA cheaper to prove, massively reduce the cost for proofs.
Testudo is a proof system designed to be cheap to generates proofs on with universal trusted setup. The goal is to have 4-20x faster to generate proofs.
Proof of Space
Find a replacement for Filecoin current proof of space with faster retrieval and cheaper proof generation.
Private Content Routing
Improve privacy in IPFS network through private content routing.
SnapDeal unlocked over 10EiB+ of committed capacity in Filecoin by making it possible to update the date embedded in Proof of Space.
Filecoin Proofs of Space are Groth16 proofs and their rate is linear to the amount of storage proven. This poses a limit to the network growth. By aggregating Groth16 proofs with SnarkPack, we removed this limit allowing for superlinear scaling.
Find a replacement for Merkle trees that is proof-friendly so that it can be used for (1) better proof of space, (2) a provable Filecoin state that can be used for computing over data and for exporting Filecoin to other chains.
Retriev provides a cryptoeconomic guarantee of retrieval. Storage providers can make a retriev deal by putting down collateral that will be slashed in case they failed at serving the file to a user and a third party committee.
Medusa: Programmable Access Control for Data
Create a decentralized access control network so that data stored on IPFS and Filecoin can have privacy functionalities.
Storage Providers run their mining book-keeping on-chain (reporting faults, proving storage, calculating power, etc.). The Neutron proposal shows how it can be done offchain using SNARKs
If Web3 is to replace Web2 then it needs to be able to handle large datasets and queries on those. Can we have verifiable Databases for Web 3?
Verifiable Computation over large committed vectors
Areas of work
Research and design protocols that improves the quality of storage services provided by Filecoin, unlock data uses in Filecoin and export Filecoin to other chains.
Maximise Data Utility
Further the utility of data onboarded onto Filecoin and the onboarding process itself by focused changes to protocols used by Filecoin ecosystem.