Software Engineer for IBM Db2 Big SQL. Lead engineer for adding support for Apache Iceberg table functionality as well as integrating with other Open Data Format tables, while storing metadata within a standalone Hive Metastore. Implemented and improved vectorized read performance of Iceberg tables by 200% comparing metrics of existing infrastructure within DB2 using 1TB TPC-DS data queries.Integrated Apache Iceberg schema evolution strategy to be leveraged by IBM’s BigSQL/Entrepot product, allowing for internal and third-party software to interpret and manage the schema. Implemented Iceberg clustered data writer support over a multi-node cluster improving overall write performance for IBM BigSQL product by 30%. Constructed metadata locking mechanism for Iceberg concurrent write operations for partitioned and non-partition tables with consideration of custom client catalogs. Architected and developed vectorized read mechanisms for Avro tables including complex and logical types

Predominantly participated with incoming clients, evaluating their specific requirements and translating their request into functional requirements for development.Primary contributor to Scala/Spark in-house configuration driven ETL and feature generation pipeline utilizing Hadoop ecosystem. Allowed client specific customization using JSON parsing. • US Patent No. US-11636391-B2 – April 25, 2023 • US Patent No. US-20220058341-A1 – Feb 24, 2022Architected and implemented search and extraction algorithms for efficient NoSQL table extraction from Apache HBase • US Patent No. US-11663177-B2 – May 30, 2023Implemented schema collection of customer data during ETL. Process captured schema information from incoming files to be stored for later use during data exploration or data extraction from Apache HBase. • US Patent No. US-20220171747-A1 June 02, 2022Created and implemented spark processes for incremental data updates during ETL. This process allowed for partial updates to be applied to flat files stored within HDFS. • US Patent No. US-20210342333-A1 – April 11, 2023Ingested customer provided flat data files to HBase tables with schema normalization.Primary contributor to client specific Apache spark tuning, determined by size of Hadoop ecosystem and client workload.Implemented a strategy for client specific database normalization. Allowing for multiple clients to have the ability to leverage the same database model