Skip to content
24. Dhyey Mavani: Applied computational mathematician and engineer episode artwork

Jul 30, 2025 29 min

24. Dhyey Mavani: Applied computational mathematician and engineer

from UserExperience.org Hosted by Olivier

Dhyey Mavani is a Software Engineer at LinkedIn, where he works on inference optimization for generative AI models within the AI Platforms org. With a rare triple major in Computer Science, Mathematics, and Statistics from Amherst College, Dhyey graduated with two summa cum laude honors theses: one in Mathematics, where he formalized chip-firing games and the Graphical Riemann-Roch theorem using Lean4, and another in Statistics, where he developed a Python package for discrete copula modeling to analyze complex ordinal response data. At LinkedIn, he built and productionized the company’s first PyTorch-Rust model serving engine, reducing GPU costs by ~$1M and powering key recommendation systems. His work stands at the intersection of applied computational mathematics, scalable infrastructure, and human-centered AI—balancing latency, reliability, and ethics. Dhyey is also passionate about EdTech innovation, open-source collaboration, and building systems that are as interpretable as they are efficient—always guided by a belief that great user experiences start with rigorous design and deep empathy. You can find more about Dhyey at linkedin.com/in/dhyey-mavani ______ If you, or someone you know, would like to be a guest on the podcast send an email to [email protected].

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

24. Dhyey Mavani: Applied computational mathematician and engineer

0:00 29:27

No transcript yet.

Requests help us decide what to transcribe next.

Frequently asked questions

How long is this episode of UserExperience.org?

This episode is 29 minutes long.

When was this UserExperience.org episode published?

This episode was published on July 30, 2025.

Can I download this UserExperience.org episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.