Riad Wahby

Most people in crypto can do one of three things well: write the academic spec, prove the underlying theorem, or ship the product. Riad Wahby keeps insisting on doing all three. His username, scattered across GitHub, Keybase and LinkedIn, is "kwantam." His company is Cubist. His refusal is the interesting part.

Today, Wahby is co-founder and CEO of Cubist, a key management infrastructure company funded by Polychain Capital. He is also an assistant professor of Electrical and Computer Engineering at Carnegie Mellon University. The two roles are not a hedge. They are the same argument made twice - that the gap between what cryptographers can prove in a paper and what developers can actually use in production is too wide, and that someone should close it.

Cubist's flagship product, CubeSigner, signs blockchain transactions inside secure hardware in milliseconds, without ever exposing the secret keys to whatever software is asking. Validators stake, bridges move assets, wallets sign - and the keys stay sealed inside hardware security modules and trusted execution environments. The pitch is almost stubborn in its plainness: "CubeSigner doesn't force teams to choose between security, performance, and convenience."

"How can we build trustworthy chips at untrusted fabs?" - the research question that follows him from the lab into the company.

That question is not rhetorical for Wahby. Before he was a cryptographer he was a circuit designer, and he knows exactly how hardware gets made: in factories you do not own, by processes you cannot fully inspect. The academic version of the problem became "Verifiable ASICs," a paper that won a Distinguished Student Paper Award at IEEE Security & Privacy in 2016. The commercial version became a startup that guards keys for Web3 teams who cannot afford to be the next headline.

The cryptography credentials are unusually deep for a founder. Wahby is a co-inventor of three proof systems - Lasso, Hyrax, and Brakedown - that show up in the toolkits of researchers building zero-knowledge applications. He co-authored RFC 9380, the standard for hashing to elliptic curves, and contributed to the BLS signature standard used by Ethereum, Avalanche, and a long list of other chains. When those networks agree on who signed what, they are leaning on math Wahby helped write down.

His path to that math was not direct. Wahby earned his SB and MEng in Electrical Engineering and Computer Science at MIT, doing his master's work in the LEES lab under David Perreault. Then he spent roughly a decade at Silicon Labs as a staff design engineer building analog and mixed-signal integrated circuits - the deeply physical, deeply unglamorous work of making real silicon behave. It is not the resume of someone who wandered into hardware security. It is the resume of someone who lived there first.

The turn toward research came through Mike Walfish, with whom Wahby worked as a visiting researcher at UT Austin and then as a junior research scientist at NYU. From there he went to Stanford for a PhD in Computer Science, advised by Dan Boneh and Keith Winstein, supported by a Ripple Fellowship. His 2021 dissertation - "Concretely efficient interactive proofs and their applications" - is exactly the bridge his career keeps building: not proofs that work in theory, but proofs efficient enough to actually run. Along the way he picked up a Best Paper Award at USENIX ATC 2018 for "Pantheon," a training ground for Internet congestion-control research.

Then came Algorand, where Wahby worked as a cryptographic researcher, and then, in 2022, Cubist. He did not found it alone. The co-founding team - Ann Stefan, Fraser Brown, and Deian Stefan - brought a track record in systems security that includes hardening Firefox and surfacing vulnerabilities in Chrome. Rather than treating security as a feature bolted on at the end, they built CubeSigner from methods that academics actually trust: information flow control, formal verification, compartmentalization, and what the company calls "intentionally simple cryptography."

In March 2023, Cubist raised a $7 million seed round led by Polychain Capital, with dao5, Amplify Partners, Polygon, Blizzard, and Axelar joining. The product launched in stages through that year: a hardware-backed key management platform in April, then CubeSigner's wallet-as-a-service for millisecond-latency remote signing in November. The framing in the press was telling - "led by computer science professors" was the headline, not "led by serial entrepreneurs." Cubist is what happens when researchers decide the paper was not the finish line.

What makes Wahby worth watching is the insistence that rigor scales. Plenty of founders promise security. Fewer can write the standard, prove the proof, and then explain - in a podcast, in a classroom, in an SDK - why a programmable policy engine beats a hope and a prayer. He still teaches. He still publishes. He just also happens to be running a company that turns the theorems into something a developer can call from an API.