Practitioner Blogs
Zalando Applied Scientist. Weekly posts on causal inference with code — CUPED, causal forests, AIPW, and more.
Amazon/Netflix economist. Causal inference notes, pyfixest maintainer, clean Python implementations of econometric methods.
'The Effect' author. Animated causal inference plots, Library of Statistical Techniques founder.
'Causal Inference: The Mixtape' author. Substack on causal inference, Mixtape Sessions workshops.
Columbia statistician. 20 years of blogging on Bayesian stats, causal inference, and social science methodology.
Microsoft economist. Demand estimation, cloud pricing, Julia packages for IO.
Principal Applied Scientist at Amazon. RecSys, LLMs, and ML systems. Author of applied-llms.org and applyingml.com. Prolific writer on building ML products at scale.
Yale economist. 'A Causal Affair' substack with Reader's Digest breakdowns of Econometrica papers. Accessible yet rigorous causal inference.
Ex-Lyft/Facebook DS lead. Co-creator of Prophet. Writes on causal inference in tech, experimentation, and 'when do we actually need causal inference?'
Oxford econometrician. Frank, educational posts on statistical pitfalls, instrumental variables, and how to read econometrics papers.
Master of R/Tidyverse explainers. Demystifying series on ATE vs ATT, DAGs, Bayesian priors. Clear code examples for causal inference.
World Bank economist. Legendary methodology posts on randomization, power calculations, and missing data — directly applicable to A/B testing.
'Metrics Monday archive. Applied econometrics advice on standard errors, weak instruments, and practical research design.
'Causal Inference for the Brave and True' author. The primary bridge between econometric theory and Python production code. DML, uplift modeling, and synthetic controls implemented.
Causal Discovery specialist. When you need to discover the causal graph from data (not assume it). DoWhy, CausalPy, and Bayesian networks.
Applied Scientist. Bayesian stats, PyMC, and geo-experimentation. Solves business problems (marketing attribution, geo-lift) with heavy-duty Bayesian math.
Co-author 'Trustworthy Online Controlled Experiments'. Deep A/B testing statistics — variance reduction, sequential testing, stopping rules without p-hacking.
'Interpretable Machine Learning' author. Opens black-box models to understand feature importance (Shapley values) — the ML equivalent of regression coefficients.
Interconnects. The primary chronicler of RLHF. RLHF as mechanism design for AI: eliciting preferences and training reward models.
Neural nets from scratch, LLMs as operating systems. Explains the intuition of optimization landscapes, not just how to write code.
Lil'Log. OpenAI Safety Systems lead. The single best technical summaries on DL, agents, and RL. Often cited as primary sources in papers.
Ahead of AI. PyTorch, LLM training, efficient finetuning (LoRA). Bridges academic papers and 'will this run on my GPU?' practicality.
DeepMind. The definitive resource on bandit algorithms. Essential for pricing, dynamic allocation, and experimentation.
Online Communities
Large Slack community (74,000+ members) for data scientists, ML engineers, and analysts. Free courses, weekly events, and active discussions on ML and data engineering.
Slack community (66,000+ members) for analytics engineers. Discussions on data transformation, SQL best practices, and modern data stack. Global meetup groups.
Discord community (3,500+ members) run by economics professors and practitioners. Discussions on economic research, econometrics, and policy.
Slack community (8,000+ members) for analytics leaders. Exchange experiences and advice on building data teams, analytics strategy, and career growth.
Slack community for data science learners. Weekly book clubs, TidyTuesday projects, and peer support for R, Python, and statistics.
Discord server (8,400+ members) for discussing econ grad school paths, life in PhD programs, and academic economics careers.