If you're a Senior, Staff, or Principal Data Scientist, you own models end to end — design, training, validation, and production. You can defend a validation strategy to a peer who will try to break it, and explain the result to an executive without hiding behind metrics. You mentor. You don't wait for someone else to make the work real.
If you lead analytics, you are not running a service desk. You set the strategy, build the team, and move the company from reactive reporting to insight that changes decisions. You zoom out to company strategy and zoom in to the pipeline and the spreadsheet. You hold the line on metric definitions, experimentation, and quality when the business wants the number yesterday.
If you build ML platforms, you own the path from experiment to production — pipelines, observability, drift, registries, infrastructure-as-code. You make data scientists faster without lowering the bar. You've run production ML on cloud infrastructure, and you treat reliability as part of the model, not a follow-up project.
At every level
- Deep technical fluency in Python and SQL, and enough cloud and data-platform literacy to ship, not just present
- Experience in fintech, lending, payments, or another regulated environment where a wrong number has consequences
- A track record of taking things from working to reliable to growing — through ambiguity, not around it
- Low ego, high standards: you do the unglamorous work and hold others to the same