• Conducted knowledge graph development in Java to express key-value pair triples as RDF data for a proprietary database ontology using Grafo and SPARQL.

• Increased runtime speed of the Java-based SQLServer metadata collector to ensure accurate retrieval of extended properties, and improved metadata collection speeds by 50% by optimizing the IRI hashing method.

• Implemented a mutex-lock-based system on the Snowflake DB, resulting in significant optimization of the integration test suite using exponential backoff, atomic transactions, and JUnit5 for parallel testing.

• Collaborated with the IT team to implement Kerberos-based authentication for the SQLServer metadata collector, ensuring secure and reliable access during the metadata collection process.

• Developed endpoints in Java and TypeScript for SPARQL query retrieval for user resources.

• Built components in Typescript React for interactive custom resource tooltips to enhance UX for metadata lineage.

• Expanded data lineage package in TypeScript to help visualize data transformations in ETL pipelines.

• Built an updated operating system version for the in-house location mapping sensor to resolve existing cybersecurity vulnerabilities (CVEs) using BalenaOS and Yocto

• Developed a nightly job to stream data from multiple SQL tables to Delta Tables in Databricks using PySpark

• Refactored existing Linux kernel modules for LED triggers and Wi-Fi drivers to ensure module compatibility with an upgraded RaspberryPi Linux Kernel.

• Scripted an image pre-processing pipeline in Flask using Tesseract-OCR and OpenCV to ensure 95% character detection accuracy in parsing bank statement PDFs.

• Developed a full-stack web app to allow users to upload, classify, and view corporate bank statements using React.

• Created a Flask API with OCR capabilities to handle CRUD operations on user-uploaded bank statements, persisting relevant data into a PostgreSQL database.