|
Suraj Reddy I am a student at MIT studying electrical engineering, computer science, and mathematics. My work sits across robotics, embodied AI, reasoning systems, and computational materials research. I currently do robotics research with the Improbable AI Lab. I previously worked on agent memory and planning through MINDSTORES, and researched computational molecular dynamics and materials research at Stanford and the University of Delaware for ~4 years. Outside the lab, I have worked as an AI/ML engineer at Amazon, worked at NASA, worked at a humanoid startup, and built projects around earthquake-resistant metamaterials, research tooling, and aerospace education. |
|
Selected ResearchI am interested in embodied intelligence, robotics, machine learning, world modeling, planning, and scientific computing. |
|
Embodied AI
MINDSTORES
|
MINDSTORES: Experience-Augmented Planning for Embodied Agents
Suraj Reddy and collaborators Research project website A framework for embodied agents to build and reuse structured memory from environmental interaction. |
|
Reasoning
ACO-ToT
|
Pheromone-based Learning: Ant Colony Optimization for Tree of Thought Reasoning
Co-author arXiv, 2025 arXiv A biologically inspired search method that improves how language models explore and select reasoning paths. |
|
Materials
Stanford
|
Molecular Dynamics Research for Solar-Relevant Nanocomposites
Suraj Reddy Stanford University paper First-authored work on PMMA-etOCS nanocomposites and durability mechanisms relevant to energy systems. |
|
Robotics
Improbable AI
|
Robotics Research at the Improbable AI Lab
MIT CSAIL Current lab Ongoing work on robotics and embodied intelligence for agents that can learn more structured, adaptive behavior. |
Coursework
6.7960 (G): Deep Learning
|
MoreA fuller archive of research, industry work, and leadership projects lives on the work page. |
| Source design adapted closely from the leonidk.github.io template, itself based on Jon Barron's website. |