Ajay Narayanan Sridhar

I’m a graduate student at the Georgia Institute of Technology, pursuing a Ph.D. in Electrical and Computer Engineering under the guidance of Prof. Vijaykrishnan Narayanan. I expect to graduate with a Ph.D. in 2027 and am open to job opportunities in machine learning and AI.

In summer 2026, I interned as a Machine Learning Engineer at Dropbox, where I worked on multimodal video retrieval for image- and text-based search. During the internship, my team and I also built an agentic AI system for customer-issue triage and diagnostics, which won the Best Technical Execution and Most Likely to Ship awards at the Dropbox Emerging Talent Hackathon.

I completed my M.S. in Computer Science and Engineering at The Pennsylvania State University in 2024. Before starting graduate school, I was a software engineer at Navis for two years (2020–2022), where I helped design and build scalable machine learning platforms for predictive and prescriptive analytics in global shipping terminal operations. My work involved developing data pipelines and streaming services using Apache Kafka to support real-time analytics and model deployment. I completed my Bachelor’s in Computer Engineering (Honors, specialization in AI & ML) from the Indian Institute of Information Technology, Design and Manufacturing, Kancheepuram (IIITDM Kancheepuram) in 2020.

Outside of research, I enjoy cycling, working out, and experimenting with creative accessibility technologies in my spare time.

research focus

My research focuses on human-centered and multimodal AI, spanning visual accessibility, AR-based interaction, and efficient large-model systems. I am currently working on developing AI and AR-powered assistive systems that help blind and low-vision users find and interact with objects in their environment using LiDAR-equipped iPhones, ARKit, and vision-language models (VLMs/LLMs). This line of work, supported by the National Science Foundation (Grant No. 2318101), explores how multimodal and generative AI can enhance independence and accessibility for people with disabilities.

I am also interested in multimodal retrieval and memory, model optimization, and adaptive inference, including contextual bandit–based routing for large language models to improve efficiency and responsiveness. My recent work explores retrieval-augmented systems for contextual recall and efficient on-device visual question answering. Broadly, my interests lie in computer vision, agentic AI, model optimization, and machine learning systems for accessibility and inclusion.