rohit.yarabati / infra

Rohit Yarabati

Building the backend for AI.

I don't build the interfaces; I build the engines. My focus spans the entire data lifecycle—from engineering robust data pipelines to training and deploying machine learning models that execute efficiently in production.

IncomingMS Data Science — Stony Brook University
B.Tech CSEAmrita Vishwa Vidyapeetham · CGPA 8.01/10
FocusLLM Infra · Distributed Systems · Graph & Vector Retrieval
rag_inference_pipeline

Projects

03 production architectures

Peer-Reviewed Research

04 publications

Technical Competency

categorized
languages
Python [core]Java [core]C++ [sys]SQL [db]
frameworks
PyTorch [ml]Scikit-learn [ml]PySpark [dist]Spark_StreamGraphFrames
infra / cloud
Docker [env]Git [vcs]OpenMPI [hpc]AWS_TextractAWS_BedrockTwilio_APILinux [os]LaTeX

About

profile
academic matrix

Rohit Yarabati (legal name Yarabati Sai Santhosh Rohit) Incoming MS Data Science candidate at Stony Brook University, holding a B.Tech in Computer Science from Amrita Vishwa Vidyapeetham (CGPA: 8.01). My work focuses strictly on the architecture beneath AI systems. Rather than building the application layer, I engineer the engines—designing persistent memory for autonomous agents, optimizing distributed streaming graph analytics, and structuring cluster-scale LLM inference pipelines."

system flow — hover a node
ingestion processing storage output

Deep-Dive System Description