At People Tech Revolution, I engineered a production RAG pipeline with sub-200ms hybrid retrieval and improved a voice assistant’s time-to-first-token from about four seconds to under one. I also developed a provider-abstraction layer for LLM and speech vendors.
I built CI-gated LLM evaluation and safety checks, and authored Terraform infrastructure for GCP services. My work also included an ETL pipeline for clinical and youth-support sources with Australian data residency controls.
At The University of Queensland, I delivered tutorials on distributed data processing and mentored students through big data analytics projects using cluster infrastructure.
Earlier, at Shell India Markets Private Limited, I automated regression testing and validated API and reporting data. In my projects, I developed and benchmarked multimodal retrieval agents and packaged a galaxy-classification model for inference.
