Madhav Kanda
My work focuses on language models, probabilistic programming, and multimodal reasoning. I am interested in making AI systems more reliable and efficient.
I studied computer science at the University of Illinois Urbana-Champaign (M.S., thesis track, 2024–2026) and IIT Gandhinagar (B.Tech. with Honours, 2020–2024). My research experience includes Microsoft, Aalto University, IISc Bangalore, and Spyne.AI.
I am a Senior Associate, Applied Research at Capital One Science in San Jose, working on LLM inference optimization, diffusion-based speculative decoding, and draft-model training.
News
Scroll for older updates ↓
- Aug 2026EMNLP 2026 — Workflow perturbation work with Microsoft Research accepted to the Industry Track!
- May 2026Capital One — Joining the Applied Research team as a Senior Associate, Applied Research, working on LLM inference optimization!
- Apr 2026ICLR 2026 — Presented our work as an oral talk in Brazil: RefineStat: Efficient Exploration for Probabilistic Program Synthesis
- Feb 2026UIUC CS News — Featured in the Siebel School of Computing and Data Science news: University of Illinois News Notes
- Feb 2026ICLR 2026 — Oral presentation (top ~1–2% of submissions): RefineStat: Efficient Exploration for Probabilistic Program Synthesis
- Jan 2026ICLR 2026 — Paper accepted: RefineStat: Efficient Exploration for Probabilistic Program Synthesis
- Jan 2026ICLR 2026 — Paper accepted: Constructive Distortion: Improving MLLMs with Attention-Guided Image Warping
- Jan 2026POPL 2026 — Presented a tutorial on constraining LLM generation syntactically and semantically, and a LAFI workshop paper (Rennes, France).
- Jan 2026EACL 2026 — Paper accepted at EACL Findings: Compositional Reasoning via Joint Image and Language Decomposition
- Nov 2025POPL 2026 — Paper accepted at LAFI: Grammar-Constrained LLM Generation for Reliable and Efficient Probabilistic Program Synthesis
- Aug 2025Completed my summer internship at Microsoft. Thanks to my mentors Alok Kumbhare, Pedro Las-Casas, Rodrigo Fonseca, and Sharad Agarwal.
- May 2025Selected for the Y Combinator's AI Startup School from among thousands of applicants.
- May 2025Started my Internship with Microsoft & Azure Research Systems, Redmond!
- Aug 2024Excited to start my second chapter as a Fully Funded MS CS student at UIUC!
- Jun 2024Awarded the Institute Gold Medal at IIT Gandhinagar for Outstanding Performance.
- Feb 2024Received an offer for the Caltech SURF program from Prof. Tapio Schneider in the CliMA group.
- Feb 2024Among the two students nominated by IIT Gandhinagar for the prestigious Pre-Doctoral Research Assistant Program at Microsoft Research India.
- Jan 2024Selected for Google Research Week 2024—exploring cutting-edge ML techniques with top experts.
- Oct 2023NeurIPS 2023 — Paper accepted at ReALML: Towards Scalable Identification of Brick Kilns from Satellite Imagery with Active Learning
- Sep 2023Selected for Amazon ML Summer School!
- May 2023Started my summer research internship at Aalto University in Finland.
- Mar 2023Among the 50 applicants selected for the Aalto University research program out of 1,600+ applicants across 84 countries.
- Sep 2022Awarded Academic Research Ranking 2 in the Computer Science and Engineering department at IIT Gandhinagar.
- Aug 2022Mentioned as a contributor in Dr. Kevin Murphy‘s Probabilistic Machine Learning: Advanced Topics book
- May 2022Started my summer research internship at Indian Institute of Science, Bengaluru.
- Apr 2022Awarded Indian Academy of Sciences Summer Research Fellowship.
- Aug 2021Awarded a branch change to the Computer Science and Engineering department for securing fourth rank among all BTech 2020 students.
- Feb 2021Secured Second Place out of eleven teams in Indusdev hackathon, a 42-hour long industrial hackathon held at IITGN