Shadow
I'm an applied AI researcher and engineer. I currently work at CoralOS as a member of technical staff. I specialise in harness engineering for multi-agent systems: the memory, context, tools, verification loops, specialist contracts, and adaptive infrastructure that let frontier models compound capability over time.
Research
Working notes, essays, and implementation-facing research on agent systems.
Amplification Architecture
A working frame for treating multi-agent systems as capability amplifiers around strong models, rather than as compensation layers around weak ones.
The Agentic Memory Iceberg
Memory becomes organizational context in useful agent systems: histories, procedures, permissions, preferences, and audit trails.
Multi-Agent Harness Engineering and Design
A frame for applying harness engineering principles to multi-agent systems: state, tools, memory, handoffs, evaluation, and maintenance as models improve.