Hey, I’m Manish.
I am a Senior Manager of Machine Learning at Zeta Global, where I focus on the infrastructure, architecture, and system design that powers our machine learning efforts. Lately, my time is spent in the weeds of production MLOps,building asynchronous distributed systems and diving deep into LLM/SLM internals, inference optimization, and memory management.
My path here started with a Master’s in Mathematics and Computing from IIT Kharagpur, followed by a foundational stint at @WalmartLabs building item-level demand forecasting models for US stores. Over the years, my toolkit has transitioned from early big data ecosystems like Hive and Spark to modern, high-throughput setups driven by Python, FastAPI, Docker, Kubernetes, and Qdrant.
I’m big on clean engineering foundations, scaling reliable pipelines, and figuring out how to make complex models run efficiently in production.