Santona Tuli’s data career started at one of the most extreme scales imaginable: fundamental physics. As a researcher at CERN, she sifted through massive event data from particle collisions to detect rare particles, work that demanded rigorous statistics, careful engineering, and a healthy respect for noisy data. From there she extended her machine learning engineering into natural language processing before shifting focus to product and data engineering for data workflow authoring frameworks, the tools that data teams rely on to define and run their pipelines.

As a Python engineer, Santona worked on Airflow, the widely used programmatic data orchestration tool, where she helped improve its usability for data science and machine learning pipelines. That experience with orchestration led her naturally into questions of how pipelines should be expressed in the first place.

Currently at Upsolver, she leads data engineering and science, driving research for the company’s declarative workflow authoring framework in SQL. Across every role, Santona has pursued the same theme: building, and empowering others to build, end-to-end data and ML pipelines that scale, whether that means a physics experiment, an open-source orchestrator, or a modern data platform.