PROJECTS
A selection of projects — some public, some ongoing. Each represents a problem I found interesting enough to think about carefully.
Software Considerations
A blog exploring the strategic side of software — architectural decisions, AI strategy, and the organizational dynamics behind technical systems. Features experiments in multi-AI dialogue as a format for exploring complex ideas. "Considerations of and by software."
softwareconsiderations.com →LGTM Agent Flight Control
A reference architecture for governing AI agent fleets: agents file flight plans before execution, a control plane enforces tiered access boundaries, and an observability layer builds behavior profiles — all on the LGTM stack and OpenTelemetry. Published as an open, versioned spec with conformance invariants and the decision records behind the design.
github.com/SiegristJ/lgtm-agent-flight-control →llmhist - Coming soon
A tool for extracting and normalizing conversations across multiple LLM providers into a consistent markdown format. Designed to preserve the substance of AI dialogues — including multi-model conversations — for reference, research, and eventual publication.
More to come. Want to discuss a project or collaborate? Get in touch.