A Day in the Life of a Data Scientist in 2026
| Source: Towards Data Science
Tags: Claude, Claude Projects, data science, prompt engineering, LLM workflows
A data scientist documents how daily work shifted in two years from manual SQL/Python to prompt engineering and LLM output validation — noting that Claude with Projects finally made AI useful for complex data work, but the job became different, not easier.
Details
Haden Pelletier at Towards Data Science describes a two-year transformation in data science workflows. In 2024 he was writing SQL and Python line by line and debugging for hours; now a meaningful chunk of his day goes into crafting precise prompts, validating LLM outputs, and managing token costs across providers. The central shift is from 2024-era ChatGPT (frustrating, lacked persistent context) to Claude with Projects and Skills, which the author credits with context persistence that made AI genuinely useful for complex, multi-session data work. Specific emerging competencies he identifies: writing constrained prompts with explicit precision requirements (decimal places, units, what not to guess), A/B testing prompt variants on the same task, and building cost optimization strategies across LLM providers. The observation that the job 'quietly turned into a different job' captures a real transition burden for practitioners retooling mid-career. The piece is candid that the change isn't uniformly positive — it documents real skill displacement alongside genuine productivity gains.