From models to operating paths
Retrieval, evaluation loops, serving reliability, and the data quality that decides whether an AI product actually works.
About · Raja Mannem
I build production AI on large data.
I'm Raja Mannem — a Senior Systems Engineer for BigData & AI/ML at Amazon Web Services. Over 11+ years I've helped enterprises run large-scale data systems on AWS and was recognized as an Amazon EMR Subject Matter Expert. Now I build the AI layer on top: retrieval, evaluation, agents, and the pipelines that keep them fed. Built at Scale is where I write it down.
What I work on.
The through-line is turning messy operational constraints into systems people can reason about — whether the payload is a nightly batch job or a model in the serving path.
Movement, shape, and trust
Lakes and lakehouses, streams, transformations, contracts, and lineage — the ergonomics that make data usable for ML.
Distributed lessons with scars
EMR, Spark, Hive, Tez, HBase, Presto, plus Kinesis, MSK, DynamoDB, Glue, and Athena — and the edge cases that only show up at scale.
Experience in brief.
A decade-plus on the AWS Big Data front line — escalations and operations on large workload clusters for enterprise customers, with open-source fixes shipped along the way.
2022 — present
Sr. Systems Engineer · BigData & AI/ML Amazon Web Services
2018 — 2022
Hadoop Systems Engineer Amazon Web Services
2014 — 2018
BigData Support Engineer Amazon Web Services
2013
Application Development Intern · 4G/LTE router Ericsson
Amazon EMR Subject Matter Expert AWSome Awards All-Star MS, Telecommunications Eng · UT Dallas
Connect.
The clearest picture of how I think is in the notes, the projects, and the reading list. To get in touch, find me here: