wen👩🏻‍💻

👋 senior data scientist. i do = data science + ai + lifelong learning with a growth mindset

[wen👩🏻‍💻 @ds_wen_]

IBM Product Manager Professional Certificate

👋 An Introduction
🤝 Foundations & Stakeholder Collaboration
📝 Initial Product Strategy and Plan
🚀 Developing and Delivering a New Product
🔄 Introduction to Agile Development and Scrum
🎓 Capstone Project
https://t.co/8sxm89CEBQ

[Samuel Wong @samuel_wong_]

How To Beat Out More Experienced Candidates For Executive Leadership Roles

https://t.co/e3DoINBkti

[Samuel Wong @samuel_wong_]

What is “loop engineering?”

There’s talk about loop engineering, but what is it exactly? I looked into it, and found triggers, cron jobs, AI slop & more. Is it a “here today, gone tomorrow” trend?

https://t.co/XJYFYbw4bO

[Samuel Wong @samuel_wong_]

Control the ideas, not the code

The 8-hour trade-off matters. Every hour spent reading code is an hour not spent on what's now the most valuable work: thinking about what the software should do, where it should go, QA, and new ideas.

https://t.co/OUhq7TH88t

[Samuel Wong @samuel_wong_]

After your next meeting or important conversation, block out five or ten minutes immediately afterward, not three days later while it's stale, but right then while it's fresh. Sit down and ask yourself one question: what could I have done better in there?

[wen👩🏻‍💻 @ds_wen_]

@ds_fafa_ 😍😍

[fafa.👩🏻‍💻 @ds_fafa_]

Don’t type, just speak

The voice-to-text AI that turns speech into clear, polished writing in every app

Available on Mac, Windows, iPhone, and Android

https://t.co/S7CuHt2eBY

[fafa.👩🏻‍💻 @ds_fafa_]

Google's AI Essentials course is the one I recommend to friends who ask "where do I start with AI?"

📚 5 courses, ~10 hours total
🤖 Covers prompting, responsible AI, productivity tools
⭐ 4.8 rating, 1.8M+ students

https://t.co/KkXyhN2Ktp #AI #MachineLearning #LearnAI

[wen👩🏻‍💻 @ds_wen_]

8 hours. That's all it takes to go from AI-curious to AI-capable. Google built the course.

📚 7 courses, beginner friendly
🤖 1M+ students, 4.8 rating
🛠️ Covers prompting, data analysis, productivity

https://t.co/ObJkHCcvcm #AI #MachineLearning #LearnAI

[Samuel Wong @samuel_wong_]

The Rosetta Stone For Business’s AI Failure Is Rewriting Every Job in Tech

https://t.co/gdO3n9qOxs

[Samuel Wong @samuel_wong_]

Your Data Warehouse Isn’t Integrated Just Because the Tables Are in One Place

Centralization is not integration

https://t.co/Ef0GfwOAm7

[wen👩🏻‍💻 @ds_wen_]

The Data Anarchy Tax: Why your team is firefighting 45% of the time

Results of the June 2026 pulse survey on organizations and data engineers

https://t.co/VMOV95YQJV

[wen👩🏻‍💻 @ds_wen_]

The Rosetta Stone For Business’s AI Failure Is Rewriting Every Job in Tech

https://t.co/A8mknvLg77

[Samuel Wong @samuel_wong_]

How Expedia Group Builds AI That Lasts at Scale

A framework for how to build, deploy, and evolve AI systems for impact and scale

https://t.co/FkGH0ci6e9

[Samuel Wong @samuel_wong_]

How to tell a data story when your audience is hostile

Most advice about data storytelling assumes a receptive audience: people who want to understand, are open to being persuaded, and will act on what they learn. But that's not always the room you're in

https://t.co/h8Uy0I6CzG

[Akshay 🚀 @akshay_pachaar]

8 RAG architectures for AI Engineers:

(explained with usage)

1) Naive RAG
- Retrieves documents purely based on vector similarity between the query embedding and stored embeddings.
- Works best for simple, fact-based queries where direct semantic matching suffices.

2) Multimodal RAG
- Handles multiple data types (text, images, audio, etc.) by embedding and retrieving across modalities.
- Ideal for cross-modal retrieval tasks like answering a text query with both text and image context.

3) HyDE (Hypothetical Document Embeddings)
- Queries are not semantically similar to documents.
- This technique generates a hypothetical answer document from the query before retrieval.
- Uses this generated document’s embedding to find more relevant real documents.

4) Corrective RAG
- Validates retrieved results by comparing them against trusted sources (e.g., web search).
- Ensures up-to-date and accurate information, filtering or correcting retrieved content before passing to the LLM.

5) Graph RAG
- Converts retrieved content into a knowledge graph to capture relationships and entities.
- Enhances reasoning by providing structured context alongside raw text to the LLM.

6) Hybrid RAG
- Combines dense vector retrieval with graph-based retrieval in a single pipeline.
- Useful when the task requires both unstructured text and structured relational data for richer answers.

7) Adaptive RAG
- Dynamically decides if a query requires a simple direct retrieval or a multi-step reasoning chain.
- Breaks complex queries into smaller sub-queries for better coverage and accuracy.

8) Agentic RAG
- Uses AI agents with planning, reasoning (ReAct, CoT), and memory to orchestrate retrieval from multiple sources.
- Best suited for complex workflows that require tool use, external APIs, or combining multiple RAG techniques.

Most architectures here involve some form of retrieval-time decision. But they all run on top of whatever was already indexed.

If that indexing step outputs messy chunks, every architecture inherits them. Improving it is a separate problem from the 8 above.

I wrote about a better unit for the indexing step. The technique:

- cuts corpus size by 40x.
- reduces tokens per query by 3x.
- improves vector search relevance by 2.3x.

And it doesn't alter the retrieval algorithm, the reranker, or the embedding model.

Read it in the article quoted below.

[jessica🩶 @ds_jessica_]

Coursera Plus. One subscription, 10,000+ courses.

📚 350+ partner institutions
🎓 Google, IBM, Meta certificates included
💰 91% positive career outcomes

https://t.co/1FMBE9nBBQ

[wen👩🏻‍💻 @ds_wen_]

8 hours. That's all it takes to go from AI-curious to AI-capable. Google built the course.

📚 7 courses, beginner friendly
🤖 1M+ students, 4.8 rating
🛠️ Covers prompting, data analysis, productivity

https://t.co/ObJkHCcvcm #AI #MachineLearning #LearnAI

[Samuel Wong @samuel_wong_]

The Context Moat Is Real, But Most Enterprises Don't Have One Yet

https://t.co/8XumuQGFVm

[Samuel Wong @samuel_wong_]

Technical skills are commoditising fast. Query-writing, dashboard-building, and report-generating are increasingly accessible to non-technical people via AI tools. If that's your entire value prop, you're competing with the tool.

https://t.co/GM5svJcZo8

[Samuel Wong @samuel_wong_]

Treat data as a core product: A model’s quality is bounded by the quality of its data. We need to maintain robust pipelines, clear lineage, reproducibility, and reusable features built with documented ownership, clear schemas, and SLAs that other teams can rely on.