Mahalakshmi Venkateswarlu

Mahalakshmi Venkateswarlu is an ML Researcher and Engineer with ~6 years of industry experience, currently pursuing an MS in Computer Science (Machine Learning specialization) at Georgia Tech’s OMSCS program.

Her research focuses on LLM inference efficiency, GPU optimization, and LLM-based agents. She is passionate about understanding and improving the systems that make large-scale AI practical from CUDA kernel optimization and GPU memory management to LLM serving infrastructure such as vLLM, PagedAttention, and Mixture-of-Experts routing. Her current work sits at the intersection of recommender systems and LLM agents, exploring context efficiency and cost-quality tradeoffs at scale.

Her paper “Less is More: Benchmarking LLM Based Recommendation Agents” was accepted for oral presentation at the LARS Workshop, ACM The Web Conference 2026 (WWW 2026)

Previously, she worked as an ML Engineer at Zoho Corporation and as a Software Engineer at Cognizant Technology Solutions, building distributed systems, cloud infrastructure on AWS, and production ML pipelines. She is also building AIwithMaha, an educational brand for making ML and AI more accessible.