🔖 Key themes and curated posts

New here? These are some topics I write & speak about. Or navigate via tags or search.

Exploring ML systems in industry and how they're implemented.

System Design for RecSys & Search: Offline vs. online, retrieval vs. ranking.
Real-time Retrieval: Examples from various companies and how to build an MVP.
Search Query Matching: Via lexical, graph, and representation learning methods.
Patterns for Personalization: Via bandits, sequences, graphs, and user embeddings.
Bandits for RecSys: Industry examples, warm-start, off-policy evaluation.
Reinforcement Learning for Recsys: Long-term rewards and explore-exploit.
Feature Stores: As a hierarchy of needs (e.g., access, serving, integrity, etc.)
Data Discovery Platforms: How they help with find data and open source options.
Bootstrapping Data Labels: With semi, active, and weakly supervised learning.

Surveys on machine learning methods.

Counterfactual Evaluation & Inverse Propensity Scoring: How to run A/B tests offline.
Measuring and Mitigating Position Bias: Accounting for bias in ordered results.
Measuring Serendipity in Recommendations: Diversity, novelty, unexpectedness, etc.
A Brief Survey of NLP: From RNN to Word2Vec to Transformer to BERT to T5.

Practices at the intersection of ML and engineering.

Challenges with ML in Production and A Practical Guide to Overcome them.
Design Patterns: Patterns in ML code & systems such as factory, decorator, proxy, etc.
ML Testing: Implementation, expected learned behavior, and evaluation metrics.
Python Project Setup: With a workflow of checks that run with each git push.

Thoughts on what an effective data science process should look like.

The First Rule of Machine Learning: Start without machine learning.
End-to-End Data Science: Why I think there should be more end-to-end visibility, cohesion, and ownership, with examples from Stitchfix and Netflix.
Scrum for Data Science: Parts of Scrum that fit the data science process well.
Agile & Data Science: What works and what doesn’t and tweaks to make it work
What I Do Before, During, and After a data science project for team productivity.

Especially in the context of a career in tech and data.

Why, What, How: A simple framework to make your writing more effective.
Design Docs for ML Systems: How to write the methodolgy and implementation sections of machine learning design docs, and the review process.
Writing for Tech Careers: Why write, my writing process, and writing at work.
Five Writing Lessons: Stuff I didn’t learn in school and my writing syllabus.

Practices that worked well for me and general advice.

Georgia Tech OMSCS FAQ: A guide to applying for and completing OMSCS.
Comparing Data/ML Roles: Data/Applied/Research scientist, and ML engineer.
Read Papers: How it helps you be a better data scientist and my approach.
Influencing without Authority: Socratic method, earning trust, finding advocates, etc.
Data Science Red Flags: What to look out for when interviewing with a new team.
Onboarding Effectively: Mindset, 100-day plan, and balancing learning and doing.
Write Weekly 15-5s: Increase visibility and earn trust with this 15-minute habit.

Random philosophical ideas and thoughts.

Summaries and permanent notes, tidied up for public consumption.

That are mostly scattered across the internet.

eugeneyan