I'm a software engineer working mainly on AI systems, backend infrastructure and Linux.
Right now I spend a lot of time evaluating models, working on retrieval, and understanding the systems around agents: context, memory, tools, state, orchestration and execution. Reliability is part of that too, especially execution safety and how models fail once they're given more capability.
I also build applications, automate parts of my environment, and experiment with hardware and connected devices.
- LLM evaluation, inference and model behavior
- retrieval and embedding evaluation
- agent and coding harnesses
- context, memory, tools and orchestration
- multi-model and long-running agent workflows
- execution safety and model failure modes
- Linux tooling and automation
- hardware and microcontroller integrations
[101, ..., 102]