Research
Active research areas. These are genuine research directions, not product roadmaps.
Automation and the structure of work
What happens to work when the repetitive, routine, and predictable elements of it can be automated reliably? What does that free up, and what does it require? This is not a productivity question — it is a structural question about how organisations and individuals operate when the drudgery is removed.
Current focus: automation depth modelling, human-in-the-loop design patterns, and the practical threshold at which automation earns its complexity cost.
Infrastructure sovereignty
The long-term consequences of centralised infrastructure dependency are becoming visible. Organisations that treated rented infrastructure as neutral have discovered it is not. This track examines what sovereign, recoverable, independently-operated infrastructure looks like at various scales.
Current focus: architecture patterns for resilient self-hosted environments, documentation standards for long-term recovery, and cost models for infrastructure ownership versus rental at different scales.
Applied AI in operational systems
Language models and AI tooling have moved from research artefact to operational component faster than the surrounding thinking has caught up. Labs explores where AI genuinely earns its place in operational systems — not as an interface novelty, but as a component that does defined work, reduces real overhead, and degrades gracefully when it fails.
Current focus: AI integration architecture for practical operational contexts, failure mode analysis, and the design of systems where AI augments without creating fragile dependency.
Robotics and physical automation
The automation of digital work is well understood. The automation of physical, repetitive, low-complexity manual work is less well addressed at the scale of small organisations and independent operations. Labs maintains an ongoing interest in accessible robotics and hardware integration.
Current focus: accessible hardware platforms, integration with self-hosted software infrastructure, and use cases where physical automation reduces burden at small operational scale.
Experimental work
Prototypes, proofs of concept, and early-stage builds. Not finished. Some will not be finished for a long time.
Lightweight framework for operational automation tasks. Minimal dependency philosophy.
Assessment of practical integration depth achievable with current LM tooling in operational contexts.
Infrastructure documentation standard for long-term human recoverability.