How AI Agents Remember Things
How do AI agents remember things between sessions? Every agent forgets everything when a conversation ends, so how do the best ones seem to know you?
I break down the memory architecture behind real AI agents, using OpenClaw (an open-source AI assistant) as a reference implementation. You’ll see how LLM agents write, store, and load persistent memory using plain markdown files, and the four mechanisms that keep context across sessions, including context window management, bootstrap loading, and pre-compaction memory flush.
Building an AI agent?
I help teams design and ship agentic systems — from architecture to production.
See how I can help
Build Your Own AI Agent from Scratch (Mastra + TypeScript)
Learn to build your own AI agent that actually does work for you, not just answers questions.
How AI Agents Search Their Memory
In my last video, I covered how AI agents store memory. But storing it is only half the problem. In this one, we dig into how agents retrieve the right memory
2. Building CreatorSignal: AI Agent Teams Build Features in Parallel (LIVE)
I used Claude Code's new Agent Teams to have multiple AI agents build features for my Rails app in parallel -- live and unedited.
How OpenClaw Works: The Architecture Behind the 'Magic'
OpenClaw agent architecture explained: How autonomous AI agents like ClawdBot create the illusion of sentience using just inputs, queues, and a loop.
Get new videos and posts by email
Weekly videos on AI engineering, plus deeper dives in the newsletter.
Occasional emails, no fluff.