Reading Notes 2025 Mar - Apr

My Medium Articles!

In the past two months, I continued writing articles per month on Medium and Towards Data Science. Here you go:

1. Google’s Data Science Agent: Can It Really Do Your Job?: Earlier this month, Google officially rolled out its 𝐃𝐚𝐭𝐚 𝐒𝐜𝐢𝐞𝐧𝐜𝐞 𝐀𝐠𝐞𝐧𝐭 to most Colab users for free. It promises to turn plain-language goals into full-on data analyses — code, visuals, models, and all. I put it to the test using a real dataset—and wrote up what it gets right, what still needs work, and what it means for the future of our field.
2. How AI Is Rewriting the Day-to-Day of Data Scientists: It is easy to get caught up chasing the latest AI tools. But at the end of the day, what matters most is using AI to eliminate what slows us down and accelerate what moves us forward. In this article, I looked at what AI tools we really need for DS and what’s out there.

Reading List in Past Two Months

Now, let’s talk about the great articles I came across in the first two months in 2025.

Data Science & Analytics

1. Does It Matter That Online Experiments Interact?: Companies usually run lots of experiments on one product at the same time. Does it matter for the results?
2. One-Tailed Vs. Two-Tailed Tests: How to decide when you should use one-tailed or two-tailed hypothesis?
3. Every single BI Tool Ever Ranked: A comprehensive list of 65(!) BI tools available in the market.
4. Statistics for People in a Hurry: Explained essential topics in statistics in easily understandable way.
5. How Facebook Sets Goals: A great walkthrough of the goal setting process at Meta and what it means for data scientists.
6. Collective Wisdom of Models: Advanced Feature Importance Techniques at Meta: Building global feature importance to improve the efficiency of feature selection and model optimization.
7. Visualizing XGBoost Parameters: A Data Scientist’s Guide To Better Models: Explains XGBoost model parameters intuitively with visualizations.

Data Career

1. Should Stakeholders be Writing SQL for Self-Service?: Why Self-service SQL might not be a good idea for non-technical stakeholders.
2. Six Organizational Models for Data Science: Six common models for data science organizations and their pros and cons.
3. The Power of Asymmetric Experiments @ Meta: When does Meta uses asymmetric experiments with real-world use cases.
4. How Data Scientists Lead and Drive Impact at Meta: How the Data Scientist role has evolved at Meta and how they drive impacts.
5. How to Recognise a Great Manager (Even If They Are Not Your Best Friend): What a great manager looks like from the five pillars of communication, trust, support, ownership, and self-care.

AI and LLM

1. From Traditional BI to GenBI: Embracing a Smarter, More Human Approach: What is GenBI (Generative Business Intelligence) and how to make it work.
2. How Facebook leverages Large Language Models to Understand User Bug Reports and Guide Fundamental Improvements: The Analytics team at Meta talks about how they utilize LLM to analyze text data like user bug reports.
3. The Impact of GenAI and Its Implications for Data Scientists: What we can learn from Claude’s analysis report on anonymized conversations.
4. Japanese-Chinese Translation with GenAI: What Works and What Doesn’t: A detailed walkthrough of the benefits and challenges of using GenAI for Japanese-Chineses translation.
5. QueryGPT – Natural Language to SQL Using Generative AI: Uber’s experimentation on text2SQL, development, iterations, use cases, and limitations.
6. Top 10 Data Science Trends for 2025: What’s Next in AI and Analytics?: How the recent development in AI would impact and reform analytics in 2025.
7. Are Data Scientists at Risk in 2025?: Whether AI will take data science jobs and how to secure your DS job from layoffs.
8. Data Science Isn’t Dying — It’s Evolving: How AI Is Reshaping the Role: What AI means for data scientists and how you can adapt to this new age.