Jay Lynch

Background

About Jay Lynch

Data engineer and educational data scientist. I spend my days building data pipelines, conducting learning science research to support student learning, and empowering university leadership with automated LLM-informed data insights. Lately, I’ve been focused on building out our university’s data engineering tech stack from scratch—envisioning the architecture, leading the team, and serving as product owner throughout. We’ve gone from nothing to a centralized, cloud-based system that brings together data from a wide range of sources using Azure Data Factory and ADLS, connected through Snowflake and dbt, and surfaced through a carefully managed Tableau environment. This work has been as much about infrastructure as it has been about impact. We've developed LLM-powered tools and Python-based apps that help leadership make data-informed decisions, detect unseen patterns, and act on emerging trends. We’ve set up pipelines and systems that enable real-time alerting, robust data validation, and ongoing monitoring—shining a light on areas that were previously data-blind. With a strong background in industry-standard languages and infrastructure (e.g, SQL, R, Python, dbt, Tableau, Snowflake, Azure, and Google Cloud Platform), I am skilled in analyzing, interpreting, and communicating complex structured and unstructured data to provide meaningful insights to stakeholders. My expertise in the learning sciences further allows me to design and support the implementation of data-driven interventions intended to drive student engagement and improve learner achievement. Past projects include building semi-supervised machine learning models to identify anomalous student learning behaviors, creating novel NLP algorithms to detect frustration and confusion in student discussions, leveraging generative AI APIs to create personalized learning tutors and provide teacher feedback, building course diversity dashboards to support educational equity, and using data to infer student learning strategies and predict course success. Additional tidbits: Published author and speaker at leading learning analytics conferences. Awarded over half-a-dozen patents in educational technology and machine learning. An unaccredited Bayesian. If interested, you can check out my blog on Medium here: