Intelligent Compression for Finance

About this session

Renç Korzay of Giza argued that decentralized finance had drifted from its founding promise of open access and was now dominated by people who knew how to read collateral positions and calculate yields. Real democratization of finance, he said, meant giving everyone access to opportunity, not simply handing out a wallet. His proposed fix was the financial agent: persistent software that manages a user's capital continuously against policies the user sets. Giza builds these agents specifically for stablecoins, which he framed as the asset class most in need of the treatment.

He described a financial agent as running deterministic, auditable logic rather than opaque models, constantly evaluating and reallocating rather than executing a single script once. To illustrate the problem, he showed how crowded the stablecoin market had become and argued that no person could track that many tokens and protocols while making fast, high-stakes decisions. Giza's answer folds all of that into one interface: a user deposits capital once, sets preferences such as which collateral or curator to trust, and the agent then acts on those preferences around the clock.

He gave several figures for Giza's own deployment, all self-reported and unverified: agents had been live for roughly eight months, had moved more than three billion dollars in what he called real, fee-bearing volume on Base rather than incentivized test activity, and had executed close to a million financial decisions. He said, again by his own account, that more than 65 percent of users customized their agent's policies rather than accepting the defaults. He also claimed, without detailing how it was measured, that users on Giza earned roughly double the yield of what he called a static or manually managed position. Separately he cited, self-reported, close to two hundred billion dollars in stablecoins sitting idle industry-wide, a gap he attributed to infrastructure lagging behind the automation now available.

He closed by naming Ethereum's security, composability and deep liquidity, along with falling L2 costs and latency, as his reasons for building there. He described Giza as already serving funds, treasuries, retail users, neobanks and wallet providers, and pointed the audience to Giza's own site to try an agent for themselves.