Part of the Omnichannel Platform Team.I utilize Splunk's Machine Learning Toolkit (MLTK) to perform anomaly detection and predictive forecasting for server data, particularly within Cassandra clusters. Focusing on the four golden signals of monitoring—latency, traffic, errors, and saturation—I develop and fine-tune machine learning models to detect anomalies in real-time and forecast potential future issues.

Part of the Technology and Business Capabilities Team. I wrote python scripts to efficiently migrate files to Azure blob storage, implemented data transfer pipelines to facilitate ETL, and leveraged Natural Language Processing (NLP) on technician help call transcripts.

Evaluated and compared three single-cell co-expression estimation methods by applying them to a public cancer RNA-seq dataset (namely: scLink, locCSN, and COTAN).

Part of Optum Tech's Enterprise Clinical Data and Interoperability (ECDI) division and serving as an analyst for the Advanced Data Acquisition (ADA) team.I improved vendor management and fueled business insights by redesigning KPIs in a submitter scorecard dashboard, developed a parsing program in Python that successfully matched 2,300 hospitals to their primary submitter, generated Power BI heat maps to highlight Optum’s national hospital connectivity, and built a k-means clustering algorithm using timeliness and report quality as dimensions to evaluate hospital performance.

Deepnote is an online notebook, similar to Jupyter, that allows teams to explore, analyze, present data and collaborate in real time within a cloud environment. Worked alongside a team of Data Science Society (DSS) consultants to predict customer churn rates by analyzing trends in user actions.