phoenix-adapters: DynamoDB on Phoenix
Your DynamoDB app, unchanged, on Apache Phoenix. No new SDK, no new data model, no rewrite.
I'm a Senior Software Engineer at Salesforce on the Phoenix Database team, where I help develop and operate a large fleet of Apache Phoenix and Apache HBase clusters. I'm interested in distributed systems, databases, and large-scale infrastructure.
Previously, I completed my Master's in Computer Science at UC San Diego, focused on distributed systems and machine learning. Before that, I was a Software Developer on the Adobe Media Optimizer team, where I built a streaming architecture for search-advertising attribution. I earned my B.Tech in Computer Science from IIT Kanpur in 2017, and spent a summer at the Adobe BigData Experience Lab integrating a topic-model-based prediction system into Adobe Illustrator.
Outside of work, I play musical instruments like drums and bass, and enjoy sports like badminton, tennis, and basketball.
Seattle, WA (San Francisco, 2021–2023)
Bengaluru, India
Bengaluru, India
Gurgaon, India
Your DynamoDB app, unchanged, on Apache Phoenix. No new SDK, no new data model, no rewrite.
Taking index maintenance off the write path with a CDC stream and an async consumer, so writes stay fast when a table has many indexes.
Why an ordered consumer needs region lineage, and how Phoenix tracks it as stream metadata in SYSTEM.CDC_STREAM.
Storing schemaless documents in a Phoenix column, and reading, filtering, and updating their fields server-side.