Earlier this month, the University of Toronto named Frank Kschischang a University Professor — the institution’s most distinguished academic rank, held by fewer than 2% of tenured faculty. Unless you work in information theory or error-correcting codes, you’ve probably never heard of him.

You’ve almost certainly used his work.

The Invisible Layer #

Kschischang co-created factor graphs, a mathematical framework for representing and computing with complex probabilistic relationships. If that sentence made your eyes glaze over, let me put it differently: factor graphs are a foundational tool used in error-correcting codes that make your wireless communications work, in the LDPC decoders inside your WiFi chipset, and in the probabilistic models that underpin a growing chunk of robotics and machine learning.

He also co-invented staircase codes for ultra-high-throughput fiber-optic transmission, which have been adopted into international communication standards. Every time data moves through a fiber-optic backbone — which is to say, every time you load a webpage, stream a video, or send a message — there’s a reasonable chance Kschischang’s work is involved in ensuring that data arrives intact.

This is the kind of engineering impact that fascinates me: completely invisible to the end user, deeply embedded in infrastructure, and operating at a scale that’s almost impossible to quantify.

Factor Graphs and the Compounding Effect #

Factor graphs emerged from Kschischang’s collaboration with Brendan Frey and Hans-Andrea Loeliger. The original 2001 paper, “Factor Graphs and the Sum-Product Algorithm,” has been cited thousands of times — but citation counts understate the influence. Factor graphs didn’t just contribute to a field; they provided a unifying language that connected previously disparate areas of research.

Before factor graphs, error-correcting codes, Bayesian inference, and signal processing each had their own mathematical formalisms. Factor graphs showed these were all instances of the same underlying computational structure. That insight unlocked cross-pollination: techniques from coding theory moved into machine learning, and vice versa.

I think about this kind of contribution a lot. Not the flashy breakthrough that makes headlines, but the quiet structural insight that makes dozens of other breakthroughs possible. Kschischang didn’t build a product; he built a lens. And through that lens, a generation of researchers saw connections they’d been missing.

Staircase Codes and Standards Adoption #

The staircase codes story is different in character but equally instructive. Co-invented with Benjamin Smith and Frank Kschischang (yes, sometimes you co-invent with yourself across papers; the naming gets recursive), staircase codes solved a specific engineering problem: how to do forward error correction on fiber-optic links at speeds exceeding 100 Gbps without introducing unacceptable latency.

The codes were adopted into ITU-T and OIF standards for optical transport networks. That adoption means they’re deployed at scale — not in a handful of installations, but across the global fiber-optic infrastructure. When you hear about undersea cables carrying terabits per second of intercontinental traffic, staircase codes are part of how that data stays clean.

Getting academic research adopted into international standards is extraordinarily difficult. It requires the work to be not just theoretically sound but practically implementable, power-efficient, and compatible with existing infrastructure. Kschischang navigated that path from theorem to standard, which is a rare combination of mathematical depth and engineering pragmatism.

What This Means for Engineering Culture #

Kschischang served as IEEE Information Theory Society president in 2010 and as Editor-in-Chief of IEEE Transactions on Information Theory. These roles reflect something beyond personal achievement; they indicate someone who invested in the health of the field itself, mentoring researchers, setting editorial standards, and shaping the direction of an entire discipline.

I find this kind of career arc particularly relevant right now. The tech industry (myself included) tends to celebrate visible impact — products shipped, users acquired, revenue generated. Kschischang’s career is a reminder that some of the most consequential engineering work is invisible by design.

Nobody opens their phone and thinks “thank God for factor graphs.” Nobody watches Netflix and appreciates staircase codes. But remove that layer of foundational work and the entire stack collapses. Every app, every service, every real-time video call depends on error correction and signal processing that traces back, in part, to the work Kschischang has spent decades refining.

The Long Arc #

I’ve been thinking about engineering impact differently since moving back to Google for RCS Business Messaging. The messaging infrastructure I work on daily sits on top of layers and layers of foundational work by people whose names most developers never encounter. Protocol designers, standards committee members, compression algorithm authors, error correction researchers — the invisible substrate.

Kschischang’s career spans nearly four decades of consistent contribution to that substrate. No pivots to startups. No viral Twitter threads. No TED talks (that I know of). Just decades of mathematically rigorous work that ended up embedded in the physical infrastructure of global communication.

There’s a version of engineering culture that would look at that career and see someone who never built a company, never scaled a product, never disrupted an industry. There’s another version that recognizes he’s had more real-world impact than most of us will achieve in a lifetime, precisely because he worked at the layer where impact compounds invisibly over decades.

I don’t think every engineer should aspire to be Frank Kschischang. The world needs people building applications and products and companies too. But I think engineering culture would benefit from taking foundational work more seriously — from recognizing that the person who makes the fiber-optic backbone reliable is creating at least as much value as the person building the app that runs on top of it.

The University of Toronto’s recognition is well-deserved. I just wish more of us knew what we were recognizing.