Build a free second brain per client with NotebookLM so Claude pulls full context before it writes a single line.
What's inside
by Divjot Sahni · EZYE Consulting · ezye.com.au
Why every new chat starts from zero
Claude has no memory across sessions, so every brief means re-explaining the same brand voice, tone, and deliverables
The hidden cost: you burn tokens and minutes re-pasting context that should already exist
The fix is a per-client knowledge base that lives outside the chat and feeds Claude on demand
Setting up your first client notebook
Open Google NotebookLM (free) and create one notebook per client
Name it clearly so you never mix client contexts
Drop in the four core source types every client memory needs...
(unlock to see the exact source list and folder structure)
Free access — enter your email below