RUNTIME LABS
About
Runtime Labs anchors language models in time, location, and context. Our work is grounded in two convictions: grounding language models in event structure increases reliability and utility of model outputs, and time is a foundational primitive for memory, planning, and coordinating between natural language intent and physical action.
We're building environments where humans and models collaborate through time, offering a shared surface for reasoning across past, present, and future.
Óra is our first platform: conversations, memory, and planning events on your personal timeline.
Research
We build systems to enable models to reason in time, keep persistent context, and work from structured representations of events. We focus on temporal indexing, memory persistence, and context retrieval so models can interpret past activity, manage future plans, and respond and schedule coherently within an evolving timeline.
Our research focuses on environments that support planning and human-agent coordination through time, where temporal understanding and shared context between human activity and models keep responses grounded and behavior reliable.
Product
Óra is built as our first integrated planning environment where models operate on persistent, time-structured state instead of isolated conversations. Context for planning is preserved as timestamped events anchored in time and location, with relevant photos and links for each event. Concretely, Óra offers version control for your schedule and conversations on a stateful timeline.
You plan with natural language and images; models interpret intent, propose events for preview and approval, and update your timeline. The system state stays continuous and aligned in time with you across past activity and forward-looking plans.
Agency
Agency emerges from the strength of our predictive models and our ability to translate intelligence into actions that shape the future distribution of decisions available to us.
Our mission is to design new tools and environments for time-aware human-computer coordination, aligning systems with human goals and strengthening agency.
Values
- We commit to clear objectives and build for the long-term, avoiding short-term optimization and cycles.
- We pursue truth rigorously and treat others with respect.
- We allocate time, capital, and compute deliberately.
- We exercise discipline and restraint in design, release, and outreach, while maintaining urgency in daily execution.
- We develop tools that empower humans—strengthening agency, judgment, and control over how work gets done.
- We adapt quickly when diminishing returns in product or research become evident.
- We maintain a respectful environment where criticism and feedback are treated as essential inputs to invention.