Systems Architecture and Agentic AI

This talk will be about the underlying concepts and ideas for building large-scale systems in the context of agentic AI!

I'll cover a few topics, such as:

• The actor model
• Tool calling, sandboxes and "code mode"
• Context Engineering and Harness Engineering
• Kubernetes and Orchestration
• Macro-agents, micro-agents and recursive language models (RLM)

By the end I hope to give a bit more insight into how agents work, and also draw some useful parallels to various other ideas in computing.