Cobbs Creek Healthcare

I built and maintained Databricks-based workflows that automated recurring processing, validation, and audit steps across 50+ healthcare/pharma datasets, replacing manual, handoff-heavy processes with repeatable jobs that run consistently, surface issues earlier, and feed downstream reporting and investigation.

I led development of a data-quality dashboard that helped teams review issues, investigate anomalies, and understand where unusual changes were happening. It organizes anomaly outputs into a usable investigation flow, from high-level issue down to the relevant accounts, territories, or aggregation levels.

I contributed to an internal agentic AI chatbot that helps users search, retrieve, summarize, and explain information across business workflows. My work focused on specific agents and tools: insurance-policy search, targeted web scraping, summarization, prompt optimization, and retrieval-backed explanation flows.

As one of three primary developers on a dementia aid patient app supporting patient and caregiver workflows, I focused on the vector database layer, database structure, and UI/UX, shaping how information is stored, retrieved, and presented in a more accessible way. A prototype/product build, not a clinically validated medical product.

Recreated overview shown above. Patient-facing screenshots are intentionally omitted to protect privacy; no real patient data is shown.

I supported patient-journey modeling using de-identified account-level data to segment behavior, support forecasting, and identify key HCPs. This involved feature engineering, clustering, and logistic regression to understand patterns across accounts for downstream business analysis.