Tiny proofs, low-memory snarks, big impact
Pratyush Mishra on the new frontiers for zero knowledge
In last week’s episode, Nico and I caught up with Pratyush Mishra, Assistant Professor of Computer and Information Science at the University of Pennsylvania. It’s been a few years since we last spoke with Pratyush, and a lot has happened since then.
Since our 2021 conversation, he has worked on projects like Garuda and Pari (tiny proof SNARKs), Arc (folding schemes based on hash functions), FICS and FACS (proofs of proximity), as well as work on low-memory snarks and zk applications beyond blockchains.
This will be our final episode of 2025, as I am taking the month of December off, but this last run of episode, including this finale, have been a great way to wrap up the year! Having Project 11, Machina, Binyi and Pratyush has meant that our final episodes touched on many of the themes and topics we have been exploring in the zk research space all year that is - cutting edge zk research, Lattices, Quantum Computing and iO!
Key takeaways from Episode 386
Pratyush has been primarily focused on hash-based and code-based systems, exploring new ways to make proofs more efficient and quantum-secure.
In Garuda and Pari, his team developed a new framework for snarks with tiny proofs, achieving some of the smallest proof sizes to date by leveraging the random oracle model.
The Pari construction in particular showed how combining polynomial IOPs with circuit-specific trusted setups can beat Groth16 in proof size — leading to ongoing work like Glock, which pushes those limits further.
Arc introduced folding based on hash functions rather than elliptic curves, providing a simpler and more post-quantum-friendly approach — though still trading off some efficiency. This is parallel to the work happening in the Lattice space.
FICS and FACS involve new proofs of proximity using “code switching” techniques, helping shrink proof sizes for linear-time encodable codes.
Pratyush and his students have been exploring low-memory snarks, like Scribe, which stores prover state on disk instead of RAM with minimal slowdown, and new time–space trade-offs for Sumcheck that dramatically reduce memory requirements.
Beyond blockchains, Pratyush and his collaborators have been applying zk tools to verifiable databases and document proofs — for example, proving that a JSON or C file follows a specific format or that a database has deleted an entry.
These projects point toward a future where zk proofs power everyday computer security: trusted data handling, proof-of-deletion under GDPR, or verifiable computation in browsers.
Was also just great to explore this part of the research space, and looking forward to seeing what Pratyush and his collaborators work on next!
Be sure to check out Episode 386: the full episode page is here, or follow this thread for commentary.
Talk soon,
Anna
ECOSYSTEM ANNOUNCEMENT
We want to hear from you!
We’re trying to learn more about what topics you’d like us to cover, where our blind spots might be, and a bit about who our listeners are. We’ve put together a short form you can fill out anonymously. We’d love to hear from you and how you think we can make the Zero Knowledge Podcast better!
Let's go Pratyush!