Scholar
Bala Thoravi Kumaravel
Google Scholar ID: RkG0-CsAAAAJ
Senior Researcher, Microsoft Research
Citations & Impact
Citations
498
H-index
10
i10-index
11
Publications
20
Co-authors
22
list available
Publications
Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Academic Achievements
- July 15, 2025: 'Grounding Task Assistance with Multimodal Cues from a Single Demonstration' accepted and presented at ACL'25 Findings.
- October 16, 2024: Presented BlendScape and SpaceBlender at UIST 2024, winning an Honorable Mention Award at UIST 2024.
- May 11, 2024: Presented SharedNeRF at CHI 2024, winning an Honorable Mention Award at CHI 2024.
Research Experience
- Before joining Microsoft, conducted research at UC Berkeley on VR/AR-assisted robotics interactions and enhancing learning experiences. During his Ph.D., he worked with collaborators at Microsoft, Adobe, and Autodesk including Cuong Nguyen, Stephen DiVerdi, Fraser Anderson, Tovi Grossman, George Fitzmaurice, and Andy Wilson. Currently, leads the development of multimodal copilots, unified natively multimodal AI copilots for Microsoft Office, generative pipelines and creative tooling for Bing Creative Ads, live AI agents for games like Minecraft, vision perception systems for AR/VR and robotics, and generative approaches to improve meeting experiences through multimodal understanding and content generation.
Education
- Ph.D. from the University of California, Berkeley, advised by Prof. Björn Hartmann; focused on Virtual and Augmented Reality with applications in diverse activities.
- Bachelor's degree from Indian Institute of Technology, Madras; bachelor's thesis won the best interdisciplinary thesis project among all engineering departments and the best thesis in the department.
Background
- Senior Researcher at Microsoft Research, Redmond in the Interactive Multimodal AI Systems group. Focuses on leveraging Generative AI models (Multimodal Large Language Models and Diffusion models) to enhance user productivity and collaboration in business-critical applications. Particularly interested in customizing, finetuning, and aligning generative AI models for specific end-user applications.
Miscellany
- Open to connecting with those exploring multimodal LLMs, diffusion models, or embodied AI for enhancing Human-AI interactions.
