Data Career Podcast: Helping You Land a Data Analyst Job FAST 201: What I ACTUALLY Do as a Data Analyst
10 snips
Mar 10, 2026 A senior analyst recounts building digital twins to simulate refinery behavior and avoid real experiments. He talks about time series smell-sensing hardware and classification models for biotech. Massive log mining and anomaly detection for cybersecurity come up, plus the practical use of Python, SQL, Excel and visualization tools across industries.
AI Snips
Chapters
Transcript
Episode notes
Prepare For Wide Role Variation
- Expect data analyst roles to vary dramatically by industry and company size, from oil giants to tiny startups.
- Avery worked at ExxonMobil, a 10-person biotech, and consults, showing tools and responsibilities change with context.
Building Digital Twins For Refineries
- Avery Smith built mathematical 'digital twin' models of oil refineries to test changes before trying them in the real plant.
- He ran hundreds to thousands of simulations, using Excel, Python, JMP, and Power BI to analyze results and inform prescriptive changes.
Prescriptive Analytics Replaces Risky Experiments
- Modeling manufacturing processes lets teams run prescriptive experiments to predict profit and risks without physical trials.
- Avery calls this prescriptive analytics and used linear and multivariate regression plus simulations to guide decisions.