The Anatomy of Memory in Humans and AI Agents
A Devoxx France conference talk connecting human memory to the practical challenges of building AI agents with memory.
Watch recordingI build and explain AI systems that work in the real world.
I’m Samuel, a developer advocate based just outside Glasgow. Through talks, courses, and notes, I help developers understand production AI agents, especially memory, context, and retrieval.
7 talks across 5 cities in 2026 · 2 LinkedIn Learning courses · 4 Code First Girls cohorts
Selected Work
About my workReduce Agent Costs with Semantic Caching
A practical LinkedIn Learning course on building semantic caching into LLM-powered agents with RedisVL to reduce repeated model calls, response time, and cost.
View courseCould agents learn to work better from their own runs?
Most research on agent memory focuses on remembering the user. I built a browser agent that learned from its own runs, and the lessons that transferred surprised me.
Read articleLatest Notes
Browse all notes-
How LLMs Turn Text into Embeddings
A first-principles look at how GPT-style language models turn raw text into token IDs, token embeddings, and positional embeddings.
-
Designing AI Agents from the Outside In
The PEAS framework offers a way to design AI agents from the outside in by defining success, environment, actions, and observations before writing code.
-
Could agents learn to work better from their own runs?
Most research on agent memory focuses on remembering the user. I built a browser agent that learned from its own runs, and the lessons that transferred surprised me.
What I write about
- AI systems agents, memory, retrieval, and what breaks after the demo.
- Life and work the pattern we’re handed, and what it takes to build a life I’d be proud of.
Learning Out Loud
New notes when I publish. No schedule, no filler.