 |
John W. Byers
AI Advisor to the Provost Professor of Computer Science Executive Director of the AI Development Accelerator (AIDA) Founding member of the Faculty of Computing & Data Sciences  Office: CDS 921 E-mail: byers @ bu . edu (Email from "me" at any other address is a scam). |
Background
- Academic Research: As Professor of Computer Science at BU, my academic
research centers on data-analytic and algorithmic challenges in two disciplines:
the science of computer networking (e.g., protocol design, network architecture), and the empirical study of Internet platforms
(such as Airbnb, Uber and Yelp). My publications are here.
- Academic Service: I currently serve as the Executive Director of the
AI Development Accelerator (AIDA) at BU, founded in
Spring 2025, and as AI Advisor to the Provost.
In 2024, I concluded a six-year term as
Senior Associate Dean of the Faculty for the College of Arts and Sciences, where I oversaw
the MCS Division of the College,
comprised of the Departments of Computer Science, Economics, Mathematics
and Statistics.
- Entrepreneurship: As founding Chief Scientist and Director at the
Cogo Labs technology incubator (founded in 2005, rebranded and operating as
Link Studio since 2025),
we developed proprietary methods for algorithmic marketing, advertising analytics, and AI tech
that powered our incubated portfolio companies from inception to profitability and beyond.
Successful exits include
EverQuote (Nasdaq: EVER),
Autotegrity and CourseAdvisor.
Cogo Labs was honored with a Greater Boston Chamber of Commerce Small Business of the Year
award in 2019.
Research Collaborations and Academic Affiliations
- BU affiliations:
Computer Science faculty,
Faculty of Computing and Data Sciences (CDS),
Networks Research Group.
-
My academic research is supported in part by grants from
the National Science Foundation. Current NSF-funded project:
- Real-Time Liquid Wireless Networking for Data-Intensive Rural Applications. Collaboration with Iowa State University and BitRipple.
- Mark Crovella and Micah Adler got me interested in working on mechanistic interpretability (understanding how LLMs tick) with them.
See our new preprint on mean field analysis of the attention mechanism in transformers.
- I've recently cycled off the ACM Internet Measurement Conference (IMC)
Steering Committee. Next stop:
Teaching
Graduate Advising
-
I've also collected a few useful
pointers to helpful advice (now very dated!) for prospective or
current graduate students.