A clearer view of AI.
THE INDEPENDENT AI.

WORTH TESTING · 19 OCTOBER 2026

The viral "Andrew Ng graph engineering course" is really DeepLearning.AI's 2024 LangGraph course

Watched. It's a genuine DeepLearning.AI short course, AI Agents in LangGraph, introduced by Andrew Ng and taught by the CEOs of LangChain and Tavily, but it was recorded in 2024 and the post's timestamps and chapters are made up. It doesn't cover 'loop engineering' or agents that rewrite themselves. Verdict: an excellent hands-on agent course if you code in Python (intermediate level, coding required); non-technical owners should watch just Andrew Ng's ten-minute intro. Top 10 to test:

Try this

  1. Swap one-shot prompts for an iterative flow: plan, research, draft, revise. 0:00:07
  2. Try the ReAct prompt loop: Thought, Action, Pause, Observation, then Answer. 0:09:41
  3. Keep the agent's memory as one running list of messages passed between steps. 0:13:59
  4. Use a simple routing rule: if the model asks for a tool, run it; otherwise finish. 0:20:14
  5. Give agents a search tool built for AI (Tavily) instead of raw web pages. 0:28:22
  6. Let the model run two tool calls at once, for example two searches. 0:27:24
  7. Add a saved checkpoint so the agent remembers each conversation thread. 0:35:34
  8. Look back through saved states to see what the agent saw when a run went wrong. 0:42:38
  9. Pause before any tool runs so a person approves it, and edit the request while paused (0:51:30). 0:46:39
  10. Stream the agent's in-between steps to your screen, not just the final answer. 0:40:17

Timestamps point to the moment in the source video.

Cost

Free to watch on X. Our notes don't cover the cost of the tools shown.

Source

Hype or Fact?

Partly fact The 'Andrew Ng graph engineering course'

How we knowSteps to try, not a claim that we independently tested every result. Follow the linked sources and check current prices. Our methods & sources ↗