I’m an M.S. student in Computer Science at Stanford University, where I also completed my B.S. in Computer Science. I conduct research with Prof. Christopher Ré in the Hazy Research Lab and Prof. Azalia Mirhoseini in the Scaling Intelligence Lab.
My research broadly focuses on understanding the fundamental bottlenecks and principles governing modern machine learning systems, and using those insights to design more efficient and capable models.
Recently, I have worked on:
Rethinking Model Design — exploring how language models represent and generate text, including Next-Concept Modeling (NCM), a coarse-to-fine framework for fast, parallel text generation.
Principled ML Systems — studying how bottlenecks in communication and computation shape system design, including an information-theoretic framework for agentic systems and Turbo-dLLM, a block-parallel training system for long-context diffusion language models.
Model Routing & Evaluation — including Smoothie, a label-free method for language model routing, and WONDERBREAD, a benchmark for evaluating multimodal foundation models on enterprise workflows.
I love exploring principled approaches to language models and ML systems, and I’m always excited to discuss new research directions and collaborations. Feel free to reach out by email!
Research Publications
Next-Concept Modeling: Fast, Parallel Coarse-to-Fine Text Generation Hermann Kumbong*, Ishan S. Khare*, Iddah Mlauzi, Andrew Woen, Charles Ding, E. Kelly Buchanan, Simran Arora†, Azalia Mirhoseini†, and Christopher Ré†. Under Review, 2026.
* Co-first author. † Equal advising. paper | blog1 | blog2 | code 
Block Parallelism for Efficient Distributed Long-Context Diffusion Language Model Training Tarun Suresh*, Pranshu Chaturvedi*, Hangoo Kang*, Parth Shroff, Ishan S. Khare, Hermann Kumbong, Azalia Mirhoseini. NeurIPS Diffusion Language Models (DiffuLM) and Long Context Foundation Models (LCFM) Workshops, 2026. paper | blog | code 
An Information Theoretic Perspective on Agentic System Design Shizhe He, Avanika Narayan, Ishan S. Khare, Scott W. Linderman, Christopher Ré, Dan Biderman. International Conference on Learning Representations (ICLR), 2026. paper | blog | code 
Smoothie: Label Free Language Model Routing Neel Guha*, Mayee F. Chen*, Trevor Chow, Ishan S. Khare, Christopher Ré. Conference on Neural Information Processing Systems (NeurIPS), 2024. paper | blog | code 
WONDERBREAD: A Benchmark for Evaluating Multimodal Foundation Models on Business Process Management Tasks Michael Wornow, Avanika Narayan, Ben Viggiano, Ishan S. Khare, Tathagat Verma, Tibor Thompson, Miguel Angel Fuentes Hernandez, Sudharsan Sundar, Chloe Trujillo, Krrish Chawla, Rongfei Lu, Justin Shen, Divya Nagaraj, Joshua Martinez, Vardhan Agrawal, Althea Hudson, Nigam H. Shah, Christopher Ré. Conference on Neural Information Processing Systems (NeurIPS), 2024. paper | website | code 
