AI Engineering.

From training data to agent context: experiments and systems behind useful models.

LEARNING PATH / AGENT SYSTEMS

Harness Engineering

Instructions, tools, state and verification around an AI agent.

LEARNING PATH / CONTEXT SYSTEMS

Memory, compression, recovery and the finite workspace of a long-lived agent.

FIELD NOTES / MODEL TRAINING

Dataset analysis, sequence length and practical training decisions.