› Protime Labs
The research streams Protime is actively working on, distinct from the consulting work on Work and the essays on Writing. Public summaries here; deeper artifacts and source are linked where available.
Also tracking: Signals — daily-curated AI agent, Microsoft 365 Copilot, and Anthropic enterprise news, scored for relevance.
Autonomous systems need more than model guardrails. AATA puts every consequential action behind identity, capability, policy, runtime, and evidence checks — including when a fleet loses its connection to central control.
Enterprise Microsoft Purview DSPM works for organizations large enough to operate it. SMB environments inherit the same shadow AI risk without the same governance budget. This research builds an edge-compute device that watches network traffic for AI service indicators and exfiltration patterns — the missing layer between endpoint controls and cloud-side data governance.
Most alignment work is constraint-based — guardrails, RLHF preference tuning, output filtering. ITP starts from a different premise: values shape what an agent considers, not just what it's permitted to say. The architecture embeds value structures into the world model and reward function so alignment happens during the agent's reasoning, not after it.
Most enterprise LLM adoption fails not at the model layer but at the integration layer — the workflows, the governance, the operational standards, the Shadow IT pattern that emerges when the central plan moves slower than the user demand. This framework is the operational shape that catches those failure modes early and makes the integration work survivable at enterprise scale.