Review an engineering plan against a curated mentor notebook
At a glance
An AI engineer uses NotebookLM with trusted technical sources and personal progress logs to challenge implementation plans. This workflow is listed under Engineering, Learning, Design Review.
- Agents / tools
- NotebookLM
SourceBuilding a tool for these workflows? See weekly rates.
What they did
1) Create one NotebookLM notebook for a specific engineering goal. Add trusted books, documentation, course notes, and conference talks, plus a one-page About Me source describing your role, goals, strengths, and recurring difficulties. 2) In Configure Notebook, choose Custom and instruct it to answer from notebook sources, acknowledge missing coverage, challenge weak reasoning, show disagreements, cite passages, and end with one action for the next day. 3) Describe your proposed agent design and ask it to compare the plan with patterns in the library. Inspect the cited passages before committing. 4) Keep weekly shipped/learned/blocked notes. Add the month's log as a source and request an assessment against the original goal, including changes for next month.
How it went
The author reports using this setup for AI engineering decisions and receiving passage-level citations to check before proceeding.
Worth knowing
The article's expensive-mentor comparison is illustrative; it does not establish equivalent coaching value or measured savings.