Hovanes Gasparian

I leverage massive datasets to build and deploy machine learning models that optimize supply chain, personalize customer experiences, and drive business growth. My work involves collaborating with cross-functional teams to deliver data-driven insights and solutions at scale.

I work on improving and expanding VIZIO's recommender systems, and personalization efforts more broadly. From data engineering in AWS, to MLOps in Databricks, I navigate complex big data pipelines and am responsible for end-to-end solutions that enhance the user experience of millions of customers.

I taught students Python for Data Science (primarily in Jupyter Lab/Notebooks). Topics range from various machine learning models, statistical methods, and algorithms for both regression and classification problems, as well as unsupervised learning methods like clustering and recommender systems. I guide students through 3 months of a very rigorous data science curriculum, including 5 projects and a final capstone for their portfolios.

I used a variety of tools to conduct data analyses for officer performance and agency insights such as SQL (Metabase and Valentina Studio) and Python (Jupyter Lab) I also engaged in natural language processing (NLP) using Google's Natural Language API for sentiment and entity analyses on unstructured data (e.g. constituent feedback and survey responses), in addition to VADER. I developed numerous dashboards in Metabase for dynamic visualizations that automated recurring analytics (e.g. monthly and quarterly statistics by agency).

I conducted various policy analyses and developed reports for presentation to Members of Armenian Parliament and the Economic Development Committee. I also collaborated with the rest of data science team to build our in-house data cluster, to speed up our data collection, cleaning, and analyses.