“We were searching for a scaling solution which would allow us to offer speed, privacy, and low-fees, all whilst ensuring auditable, secure, self-custody of assets for traders.”
Co-Founder & CEO
The system we used previously did not sufficiently allow us to serve our target audience of professional traders. The main issues were the slow speed of Ethereum Layer 1, lack of privacy of transactions, and high-cost of settlement (leading in turn to high minimum order-sizes).
It was clear that as DeFi continued to grow and the customer base became larger, that an alternative solution would be needed. We were searching for a scaling solution that would allow us to offer speed, privacy, and low-fees, all whilst ensuring auditable, secure, self-custody of assets for traders.
The main alternative scaling solution we considered was an in-house designed and built variant of ‘plasma’ that our team had developed in Rust over the course of 9 months.
We favoured a validity proof for settlement over game-theoretical proofs and security, because of their elegance, and the reduced complexity of the overall system (such as staking, roll-backs, etc).
We did not have the cryptographic expertise in our team to build neither STARKs nor SNARKs, and the development tools we would need did not exist.
We chose StarkWare due to their large team, cryptographic prowess, and belief that they would be able to produce a production-ready system faster than alternatives.
On the whole, the integration process was smooth and well managed, and we were continuously impressed by the caliber of the StarkWare team throughout. Working with an external team offered us major efficiencies, as we could then focus solely on the exchange and leave the cryptography to the experts.
We can now settle a trade at a better price than the user had signed it. This allows execution at the best price available. We are also able to decrease minimum order-sizes and keep trade history private for traders.
In theory, using Cairo our team could independently develop features that we would like based on customer feedback.