Sherry Shen

Trained strategist, product builder, lifelong learner, writer at heart. Phoenix rising, amateur coach, clumsy but loving mom.

Email · LinkedIn · GitHub

Learning AI

A path into AI for people without a computer science background. What I read and watch to tell what is worth building, while moving toward a PM or deployment strategist role at an AI company.

Technical

How Models Work

How a model is built, and what happens when you send it a prompt.

Deep Dive into LLMs like ChatGPT. Andrej Karpathy. Video, 3h 31m. Makes everything after it make sense.

Prompting & Context

What goes into the model, and what stays out.

Effective context engineering for AI agents. Anthropic. Article, 14 min. Why prompting became context engineering, and what actually earns a place in the window.

RAG & Memory

Giving a model the right information at the right time.

Complete RAG Tutorial 2026. KodeKloud. Video, 48 min. The hands-on one. You build the pipeline in the browser instead of reading about it.

Agents & Harness

How an agent plans, acts, and recovers.

A practical guide to building agents. OpenAI. Article, 17 min. Read this first. It gives you the overall map of what an agent is made of.
Building Effective Agents. Anthropic. Article, 14 min. When a workflow is enough and you do not need an agent.
You Can Learn AI Agent System Design In 19 Min. Sean‘s AI Stories. Video, 19 min. The practical one. It builds a real customer service agent instead of talking about agents in the abstract.
Harnesses in AI: A Deep Dive. Tejas Kumar, IBM. Video, 20 min. Uses a vivid real example to talk about what a harness is.

MCP & Integrations

How an agent reaches tools, data, and other systems.

Coding Agents

How coding agents work, and how their output gets checked.

Evals

Measuring quality, and reading failures to find what to fix.

Why Everyone Should Know About AI Evals. ByteByteAI. Video, 9 min. Nine minutes to get the shape of an eval before reading the long version below.
Demystifying evals for AI agents. Anthropic. Article, 30 min. The reference for agent evals. Read it before designing an eval set, not after.

Traces & Observability

Seeing what happened in production.

Feedback & Data Flywheels

Turning real usage into the next version.

Safety & Security

How these systems get attacked, and what an agent is allowed to do.

Multimodal AI

Voice, vision, video, and documents.

AI UX

What the user sees, what they can undo, and what needs their approval.

Production Architecture

Running it inside someone else’s environment.

Business

Use Cases & Workflow Design

How to tell which work is worth handing to AI, and which is not.

Identifying and scaling AI use cases. OpenAI. Article, 17 min. Where to start when nobody has told you what to build yet.

Value & Success Metrics

Whether the thing you shipped is worth what it costs.

Cost & Latency

What a request costs and how fast it returns.

Model Choice

Which model runs which job, and how requests get routed.

Pricing & Monetization

What you charge for it.

API & Developer Platform

Running the platform other people build on.

Compliance & Data Rules

What a buyer’s security and legal team ask for.

Deployment & Adoption

Why deployments stall, and what moves them.

Market & Strategy

Where the value goes.

Builder Stories

How the people who built these products talk about doing it.

The 5 Hidden Rules Behind Successful AI Products. Chris Pedregal, Granola. Video, 38 min. The short one. Start here if you only watch one.
How to Build a Beloved AI Product. Chris Pedregal, Granola. Video, 1h 09m. The same story at length, with the product decisions spelled out.
Inside the rise of Gamma, from dismissed idea to 100M ARR. Grant Lee, Gamma. Video, 1h 54m. The clearest account of finding fit by fixing the first thirty seconds.

Click to open

Sherry Shen ↓ Projects