[drafting]
We can teach subject-matter experts prompt engineering—but then what is our value-add as a data scientist?
- Technical connector. We serve as the bridge between SMEs and engineers. We build PoCs, run MVPs, and prove technical viability.
- Safeguarding CI/CD. We act as stewards of the CI/CD process, enforcing best practices that may be unfamiliar to SMEs, such as proper version control.
- Fluency in frontier research. We stay well-versed in cutting-edge research, understanding how the latest developments can help us iterate on models. This requires deep familiarity with deep learning and neural network literature.
- Disciplined model development. Evidence-driven development, edge-case handling, and robust test cases are essential to ensuring feature stability. Today, almost anyone can build a demo—but systematizing it reliably remains a core expertise.