Real-time AI and agentic systems
I work on the architecture behind reliable voice agents: speech, orchestration, evaluation, and the workflows that connect models to real businesses.
New York, NY
Christian Yongwhan Lim
I build AI voice agents at Probook, teach competitive programming at Columbia, and have spent fifteen years in the ICPC as contestant, coach, judge, and organizer.
- Chief Architect · Probook
- Adjunct, Computer Science · Columbia University
- Director of Internships · ICPC Foundation
- Formerly Google Research & Two Sigma
New York, 2026
What brings you here?
Speaking, teaching, and workshops
I speak about engineering careers, competitive programming, AI systems, and the habits that help technical people keep growing.
Competitive programming and mentorship
Course records, evaluation data, problem-solving profiles, and a public directory of the students and professionals I have mentored.
ICPC programs and technical talent
I help run internships, curriculum, conferences, judging, training camps, and teams across the global competitive-programming community.
About
I studied math and computer science at Stanford, did my master’s in computer vision with Fei-Fei Li, and started a PhD at MIT with Asu Ozdaglar before leaving for industry: six years at Google, where I led Kick Start; a stretch at Two Sigma; and startups since — first as VP of Engineering at Arklex, now as Chief Architect at Probook, where we build AI voice agents for home service businesses.
Competitive programming has been the constant through all of it. I competed for Stanford, helped coach MIT’s team during my PhD, and coach Columbia’s today — two World Finals so far. Along the way the volunteer work turned into real jobs: I teach Columbia’s competitive programming courses, run the ICPC Foundation’s internship program, chair its symposium, help edit its journal, and still solve problems myself. Josh Alman and I are writing a book about the craft.
If you’re working on real-time AI, competitive programming, or computer science education — or you’re a student figuring out your next step — write to me at yongwhan@yongwhan.io. I read everything.
Now
Building at Probook
Designing the technical architecture for AI voice agents serving home-service businesses, from real-time speech through reliable task execution.
Teaching at Columbia
Preparing the next competitive-programming and technical-interview courses while continuing to coach Columbia’s ICPC community.
Growing ICPC programs
Running the Foundation’s internship portfolio and contributing to curriculum, the CLI Symposium, the journal, judging, and training.
Writing the next book
Working with Josh Alman on a practical path from introductory algorithms to contest-level problem solving.
Working Notes
Reliable voice agents are systems problems
The model matters, but the experience is won or lost in everything around it: latency, interruption handling, tools, state, evaluation, and recovery.
Read the note →Practice that compounds
Competitive programming becomes useful far beyond contests when practice is structured around reflection, transfer, and increasingly precise feedback.
Read the note →Mentorship is leverage, not a shortcut
The best mentoring relationship does not remove hard work; it improves the quality of the questions, decisions, and feedback surrounding it.
Read the note →By the Numbers
637+
Research citations
Computer vision, networks, and programming languages — CVPR, IJCV, OOPSLA.
Google Scholar107
Completed ICPC internships
ICPC Foundation program, with 19+ active interns as of July 2026.
Program record14×
Course offerings at Columbia
Competitive programming and interview preparation, including scheduled offerings through Fall 2026.
Course record