StarkHouse is a hackathon project at MuBuenos in Argentina in April 2024.
Starkhouse leverages Giza tech, a ZKML database to provide fair and secure housing matching between potential renters and house providers on the Argentinian market. This project aims to leverage efficient and fair matching on a secure database, not accessible to any third party. This app enables the treatment of highly sensitive data. The model can be extended to any matching system on any population.
Currently, we can remark an inefficiency in housing attribution, with an unfairness of the system not based neither on the urgency nor the need.
Currently, when you're looking for to rent an apartment, you need to provide some sensitive information like the amount on your bank account, for instance.
StarkHouse MVP leveraged ZKML, which is powered by a custom machine-learning system based on an optimized linear regression. Trained on a curated Dataset of rental properties in Argentina and matched with a dataset of Argentinian credentials and revenues.
- Giza AI actions to
- Orion (Giza) to generate proofs and verify them
- Linear Regression to establish a score and match houses and users
- QuarkID to use an Argentinian ZK database
- Database of rental properties in Argentina
1 - Access to a private database (QuarkID / Rental properties)
2 - Use the algorithm to have a ranking of 5 more promising houses using linear regression for scoring per person with an extensive filter
3 - Verify the ranking is fair with verifiable Giza proof
Coming soon.
StarkHouse is a verified system providing proof of fair matching on Starknet provided by Giza. This application can power any private or public housing marketplace with fair and efficient matching.
User going to encrypt their sensitive data to rent a house as verifiable credentials.
- New use cases unblocked using any matching system with sensitive data (grants, marketplace, rewards)
- SDK to improve the matching system on any housing marketplace
- Leverage Herodotus collaboration with Giza to ensure data authenticity and provenance
- Public systems: any allocation system that matches people with needs
- Private systems: custom matching programs depending of the need