Reading Notes 2026 Jan - Feb

My Medium Articles!

Let me start with my new article – in the last two months, I published one article on TDS:

1. Your First 90 Days as a Data Scientist: In this article, I look back my latest onboarding experience and summarized practical onboarding tips for data scientists.

Reading List in Past Two Months

Now, let’s talk about the great articles I came across in the past two months.

Data Science & Analytics

1. Causal ML for the Aspiring Data Scientist: An introduction of important concepts and tools in causal ML.
2. Randomization Works in Experiments, Even Without Balance: What causes covariate imbalance even with randomization and can we still trust experiments in that case.
3. How to Model The Expected Value of Marketing Campaigns: Combine purchase likelihood models, expected value modeling and the profit curve to make marketing campaign decisions.
4. Multi-Attribute Decision Matrices, Done Right: Multi-attribute decision matrices (MADM) are a useful methodology for comparing multiple alternatives and selecting the choice that best fits your needs and budget.
5. Data Poisoning in Machine Learning: Why and How People Manipulate Training Data: Data poisoning is changing the training data used to produce a machine learning model in some way so that the model behavior is altered – how it happens, why it matters and how to avoid it.
6. The Three Ages of Data Science: When to Use Traditional Machine Learning, Deep Learning, or a LLM (Explained with One Example): Use sentiment analysis example to illustrate the appropriate use cases of traditional machine learning, deep learning, and LLM.
7. Does More Data Always Yield Better Performance?: An experiment of model performance with more samples vs. more features.
8. Time Series Isn’t Enough: How Graph Neural Networks Change Demand Forecasting: Demand forecast in a supply chain is not independent but networked, therefore graph neutral networks could be a better solution then time series models.
9. How To Estimate Correlation Between Metrics From Past A/B Tests: Data scientists at Booking.com explains how they combine two naive approaches to estimate correlation between metrics from past experiments.
10. Why Your Usual ML Metrics Don’t Work for Decision Models: A practical guide to evaluating counterfactual predictions with matching techniques.
11. The Hidden Choice Behind Every Moving Average: Explains the difference of centered moving average and trailing moving average, and their respective use cases.

Data Career

1. Why Human-Centered Data Analytics Matters More Than Ever: Always ask who is this for and how will it actually be used?
2. The Real Challenge in Data Storytelling: Getting Buy-In for Simplicity: Practical tactics to persuade stakeholders for data story-telling.
3. Is the AI and Data Job Market Dead?: Data science is not dying; it’s evolving.
4. Why Companies Stopped Hiring Junior Data Scientists? (Whom They Hire Instead?): What type of data scientists do companies need today with the advancement in AI.
5. Is Data Analytics Still a Good Career Choice in 2026?: AI is a multiplier, not a replacement.
6. 17 Rules to Mentally Survive Any Corporate Job: Helpful common career advices.

AI and LLM

1. How to Personalize Claude Code: How to get more out of Claude code by giving it access to more information.
2. Achieving 5x Agentic Coding Performance with Few-Shot Prompting: What is few-shot prompting and how to use it to improve LLMs performance.
3. How to Maximize Claude Code Effectiveness: Practical Claude Code tips, including using slash commands, memory, plan mode, etc.
4. Can AI Solve Failures in Your Supply Chain?: How to use AI agents to enhance root cause analysis in supply chain.
5. Advance Planning for AI Project Evaluation: Always have a detailed evaluation plan before starting any LLM projects.
6. How Cursor Actually Indexes Your Codebase: How Cursor uses RAG to index the codebase and provide code suggestions.
7. The Missing Curriculum: Essential Concepts For Data Scientists in the Age of AI Coding Agents: Key developer concepts, such as code smells, to enable you review the code AI generated more effectively.
8. Build Effective Internal Tooling with Claude Code: Thanks to the coding agents, the cost of building internal tooling has come down drastically. This article talks about reasons, use cases, and how to build these internal tools.