Computer Science Ph.D.
& MBA Candidate
Stanford University
& MBA Candidate
Stanford University
Advisors
Current Mentees
Hangoo Kang, Harsh Singh, Matthew Hart, Orhun Akengin, Tanvir Bhathal, Tarun Suresh
Past Mentees
Adrian Lafuente Gamarra (Salesforce), Brendan McLaughlin (Reflection AI), Herumb Shandilya (Mixed Bread), Lichu Acuña (Stealth Startup), Robby Manihani (Pace), Wes Griffin (Stanford)
About Me: I am a joint Computer Science Ph.D. and MBA student at Stanford University, advised by Azalia Mirhoseini (Scaling Intelligence Lab) and Christopher Ré (Hazy Research). I am affiliated with the Stanford NLP Group and the Stanford AI Lab (SAIL). I am also a Google Student Researcher with Sheng Li on the TPU and Gemini teams.
My research lies at the intersection of language models and ML systems. Most recently, I've studied the intelligence efficiency of LM systems, with the goal of commoditizing intelligence through increasingly efficient open-source LMs and hardware accelerators. By reducing the energy, compute, and capital required for deploying LMs at scale, we hope to make LM systems more broadly utilized around the world. Our agenda spans foundation models, ML systems, electrical engineering, and economics, and is anchored by the Intelligence per Watt project.
My doctoral studies are supported by the Stanford Graduate Fellowship, JP Morgan AI/ML Fellowship, Stanford EDGE Fellowship, and GEM Fellowship. I am a recipient of the Fulbright Scholarship (Research Award, Germany) and Gates-Cambridge Scholarship at the University of Cambridge (Trinity College) for post-graduate studies. Previously, I was a Predoctoral Young Investigator (PYI) at the Allen Institute for AI (AI2) and completed the joint B.S./M.S. in Computer Science at Georgia Tech as a Stamps President's Scholar.
My research is generously supported by Stanford HAI, Laude Institute, Lambda Labs, and IBM Research.
Selected Publications
-
OpenJarvis: Personal AI, On Personal DevicesPreprint 2026
-
Intelligence per Watt: Measuring Intelligence Efficiency of Local AIPreprint 2025
-
Weaver: Shrinking the Generation-Verification Gap with Weak VerifiersNeurIPS 2025
-
LMUnit: Fine-grained Evaluation with Natural Language Unit TestsEMNLP 2025
-
Archon: An Architecture Search Framework for Inference-Time TechniquesICML 2025ICLR 2025 Scaling Self-Improving Foundation Models without Human Supervision (SSI FM) Oral Presentation
-
ARES: An Automated Evaluation Framework for Retrieval-Augmented Generation SystemsNAACL 2024 Oral Presentation
In the News
-
Slingshots // Three: Our Third Grant CohortLaude Institute · Jun. 2026 · OpenJarvis
-
Slingshots // Two: Our Second Grant CohortLaude Institute · Feb. 2026 · Intelligence per Watt
-
How AI Will Reshape Computer Systems by 2035: A Jeffersonian Dinner in San Francisco about Our 10,000x FutureComputing Research Association (CRA) · Apr. 2026 · Intelligence per Watt
-
Scaling Token Factory Revenue and AI Efficiency by Maximizing Performance per WattNVIDIA Developer Blog · Mar. 2026 · Intelligence per Watt
-
Microsoft's Maia 200 Rewrites the Rules of Agent Intelligence per WattGartner · 2026 · Intelligence per Watt
-
Small Models, Hazy ResearchIBM Research · 2025 · Intelligence per Watt
-
Intelligence per Joule: The New Metric for True AI Value and EfficiencySambaNova Systems · Nov. 2025 · Intelligence per Watt
-
Intelligence per PicojouleReiner Pope (MATX), with Clive Chan and Dylan Patel · Apr. 2026 · Intelligence per Watt
Invited Talks
-
“OpenJarvis: Personal AI, On Personal Devices”
- AMD Advancing AI Day: Jul. 2026.
- Y Combinator (YC) Paper Club: Jul. 2026.
-
“Intelligence per Watt: Measuring Intelligence Efficiency of Local AI”
- Two Sigma: Feb. 2026.
- EE292P: Atoms, Bits, and National Interest (ABNI), Stanford University: Feb. 2026.
- MBZUAI Speech and NLP Symposium: Jan. 2026.
-
“Weaver: Shrinking the Generation-Verification Gap with Weak Verifiers”
- Measuring Intelligence Summit at PyTorch Conference: Oct. 2025.
- Together AI: Jul. 2025.
-
“Archon: An Architecture Search Framework for Inference-Time Techniques”
- LLMs Meet Data Processing Workshop, UC Berkeley: May 2025.
- Databricks: Feb. 2025.
- Contextual AI: Oct. 2024.
Fellowships & Awards
- JP Morgan AI/ML Fellowship 2025
- Stanford Graduate Fellowship 2023
- Stanford EDGE Fellowship 2023
- GEM PhD Fellowship — National GEM Consortium 2023
- Knight-Hennessy Scholarship, Finalist — Stanford University 2023
- Gates-Cambridge Scholarship — Bill & Melinda Gates Foundation 2023
- Fulbright Scholarship — Research Award, Germany 2022
- Summer Venture in Management Program — Harvard Business School 2022
- U.N. Millennium Fellowship — United Nations 2021
- D.E. Shaw Nexus Fellowship 2020
- Stamps President's Scholarship — Georgia Tech and Stamps Foundation 2018