From Psychology Degree to Data Science Lead: A Career Journey
A data science leader's account of entering the field from a non-technical background highlights the roles of persistence, failure, and luck.
AI & ML
Frontier models, research results, agents, and developer tooling.
Practical Tips for Maintaining Machine Learning in Production
A practical guide argues that maintaining machine learning models in production can be made easier with the right approach.
Why Data Science Teams Need Room to Fail
Eugene Yan argues that a tolerance for failure is essential for data science teams to innovate successfully.
Commando, Soldier, Police: A Metaphor for ML Career Choices
Eugene Yan's simple framework helps AI/ML practitioners weigh startup risk against larger-company stability.
Recommender Accuracy Isn't Everything: Serendipity Metrics Matter
A review of recommender-system research shows why surprise and discovery deserve a place alongside accuracy.
Eugene Yan Keynote: How Asia's Tech Giants Scale SuperApps
A 2019 keynote from Eugene Yan examines the scaling strategies and SuperApp approach of Asia's largest tech companies.
OMSCS AI Course Review: Simple Solutions Beat Complex Models
A detailed review of Georgia Tech's OMSCS CS6601 highlights a practical mantra—start simple, then add intelligence—that resonates beyond the classroom.
Why Rushing Into a Data Science Project Is Counterproductive
A new essay argues that deliberately slowing down before a data science project is more efficient than diving straight into model building.
Crocker's Law and the Skill of Accepting Hard Feedback
One essay explores how Crocker's Law and cognitive dissonance can turn uncomfortable feedback from a threat into a path to growth.
Spark+AI Summit 2020: Notes on Production Applications
Eugene Yan's curated notes from Spark+AI Summit 2020 focus on application-specific talks and frameworks for using Spark in production.
Execution, Not Just Planning, Drives Data Science Project Success
A data science leader argues that what happens during a project matters as much as the strategy behind it.
Why Data Scientists Should Own the Full Lifecycle, Not Just Models
Eugene Yan argues that data scientists increase their impact by taking work from problem framing to deployment and monitoring, with caveats for context.
OMSCS FAQ: Admission, Time, and Career Value From One Grad's Experience
Eugene Yan's first-person FAQ covers the practical questions about Georgia Tech's online computer science master's program.
Adding Checkboxes and Downloads to FastAPI-HTML UIs
A practical guide shows how to extend a FastAPI app's HTML front end with user-selectable options and a download button for generated results.
Automating Python Project Checks With Git Push Workflows
A practical guide shows how to run tests and checks automatically on every git push to improve collaboration and code quality.
Airflow's 'One Day Late' Jobs Are a Scheduling Semantics Issue
A new post explains why Airflow jobs can appear to run a day late—and why that's actually the platform working as intended.
Eugene Yan on Effective Data Science: Leadership, Agile, and ML in Production
A podcast episode distills Eugene Yan's practical lessons on running data science teams and shipping machine learning systems.
Automating GitHub Profile README Updates
A practical guide to keeping a GitHub profile README current with recent writing through automation.