Come be an eng manager at @convex.
The team is exceptional and we're solving genuinely hard problems designing the backend platform for agentic dev and agentic apps.
Hiring for infra and API/product roles. In-person in SF. Hit us up at our jobs page.
I am currently in West Africa teaching university students and faculty about AI and databases. It has already been quite eye-opening for me. AI could not even understand local languages until recently. Some people are worried that they don’t know how to pick projects, majors, and jobs (or advise students) now given how quickly AI is improving. Some thought we have solved these problems in the US and asked for advice. We had to say, we definitely have not solved these problems in the US.
My takeaway so far is that while we all may be at different points on the AI adoption curve, we are all similarly lost when it comes to envisioning how, in the long term, to harness AI for education, AI for science and social science, AI for building computer systems, etc.
Congratulations to @CMUDB's #1 ranked PhD student @lmwnshn for finishing! I've known Wan since he was a sophomore. He is one of the smartest + genuinely kindhearted people I've ever met.
My former #1 PhD students @lmwnshn+@17zhangw have a new startup @OneWillAI on controlling agent behavior. Git protects your code. But agents do more than write code. Databases already solved this.
OneWill make agents live database-centric lifestyles.
Demo: https://t.co/48hrCiwMHS
Unreliable systems and misbehaving agents are surprisingly similar. Odd behavior. Messy leftover state.
After graduating from @andy_pavlo's lab @CMUDB, @17zhangw and I started @OneWillAI to adapt decades of database solutions to today's agent problems.
https://t.co/RHb2EkBMIw
@charliermarsh SQLite tables are indexes. SQL Server tables are indexes. MySQL InnoDB tables are indexes. Oracle RDBMS "index organized tables" are indexes. PostgreSQL tables and Oracle default tables are heaps, not indexes.
This Fall at CMU we're teaching a new course on AI Agents!
The goal is that you learn how to create a scaffold, build evals, and train an agentic LLM using RL.
We'll try to balance theory and practice, and introduce modern frameworks and best practices.
Unreliable systems and misbehaving agents are surprisingly similar. Odd behavior. Messy leftover state.
After graduating from @andy_pavlo's lab @CMUDB, @17zhangw and I started @OneWillAI to adapt decades of database solutions to today's agent problems.
https://t.co/RHb2EkBMIw
This is slightly incorrect. The Berkeley team originally chose the process-model for Postgres in the 1980s because multi-threading would require OS-level support that was not common as it is today.
Source: The Design of Postgres (1986) https://t.co/7kGCRngfKG
PostgreSQL chose a process-based model adhering to the classic UNIX philosophy: "Do one thing and do it well."
The community decided Postgres should be a world-class SQL engine and let the mature, battle-tested operating system kernel handle block-level caching.
- then on Thurs, we will be presenting Argus led by @MangQiuyang. Argus automatically discovers new SQL testing oracles using LLMs and formal verification, enabling more effective DBMS fuzzing and uncovering 40 previously unknown bugs across major database systems.
@andy_pavlo guided us through LLM-powered improvements in database optimization.
If tuning Postgres becomes easier and easier, what remains as the real frontier?
Hint: optimizers!
Excited and deeply honored to receive the NSF CAREER Award to support our research on next-generation database systems powered by hybrid classical-quantum computing.
https://t.co/39XPGG2ose
#NSF#CAREERAward#HybridClassicalQuantumComputing#Databases#USC
One week to SAO at @CAISconf. Eight months in the making.
At VLDB 2025, we met @adityagp as he presented "Supporting Our AI Overlords" (with @shulynnliu, @matei_zaharia et al). Many ideas felt immediately familiar to us at @Bauplan_labs!
@EdwardRaffML@pvldb Yo. I just run the twitter bot. Email the editors-in-chief. They will get you in touch with your meta-reviewer / associate editor: https://t.co/fI1EDrLmVJ
I'm joining Carnegie Mellon's CS Department (and HCII by courtesy) as an assistant professor in Fall 2027!
I'll be recruiting PhD students next cycle. If you're interested in AI systems or human-AI collaboration, list me in your application. Stay tuned for more about my new lab!
I'm joining Carnegie Mellon's CS Department (and HCII by courtesy) as an assistant professor in Fall 2027!
I'll be recruiting PhD students next cycle. If you're interested in AI systems or human-AI collaboration, list me in your application. Stay tuned for more about my new lab!