- Worked in a cross-functional team developing the ML pipeline for a Epilepsy diagnosis product - Built algos to identify and score named entities from healthcare research...more papers. - First prototype achieved a ~70% recall in identifying tokens & < 10% test error in classifying into 700+ labels - Beta tests were endorsed by Mayo Clinic.
- Asia-Pacific technical support executive for an airline scheduling software suite - Owned backlog for customer requirements & issues. Managed releases of new features -...more Created and nurtured best practices for the product - Tech stack mainly SQL and shell scripts
- Implemented a predictive model to recognize critical spares in the inventory by analyzing maintenance data - Collected data to execute Reliability Centered Maintenance...more as part of a Lean initiative.
Tested and evaluated workflows and facilitated deployment of Process Control and Failure Modes & Analysis software for customers such as Bosch, Mercedes Benz, General...more Motors, Cummins, Tata Motors, etc. Was part of a team that trained shop-floor employees on software usage.
Model-building for sparse/high-dimensionality data
As a summer intern in an early-stage healthcare-info startup, I worked in a team developing a data pipeline to use NLP methods to extract healthcare information from text of research papers. Our first prototype achieved a ~70% recall in identifying tokens & ~10% test error in classifying into > 700 labels.