How The Full-Stack Data Scientist Is STILL The Sexiest Job

When MLOps first began to emerge, the promise was to increase innovation by minimizing or automating the toilsome work of data scientists. And yet, the MLOps tooling ecosystem remains fragmented. Companies just starting on their journeys to becoming ML-native or ML-fluent plan their roadmaps based on ill-fitting MLOps maturity models that don't account for their particular organizational goals or trajectory, while leaving data scientists out of the loop. Toss in the the hype and ("seemingly" widespread) adoption of LLM’s and GenAI applications as well as debates about the existence of "prompt engineers" and it would seem like the data scientist role is headed for an extinction event. My goals for this talk are to: - Illustrate the current state of the MLOps landscape; - Share my experiences from both sides of the fence as a former data scientist and a MLOps engineer working for companies like Autodesk, Teladoc, and Mailchimp; - Illustrate the common pain-points by companies around production ML; - Show how the future will favor the builders, regardless of whether they're "AI Engineers", "full-stack data scientists", or "ML Engineers".