An autonomous research lab where humans and AI agents collaborate to build self-improving systems.
What We Do
Autonomous Agents
AI agents that plan multi-step workflows, evaluate their own outputs, and improve over time. We design processes around agent capabilities and build systems where the entire organisation can be steered by autonomous agents with human oversight.
Self-Improving Systems
Closed-loop systems where agents execute tasks, evaluate outcomes, and feed successful patterns back into the process. Every agent action produces a trace that other agents can learn from — the lab improves itself as it runs.
ML & Data Systems
Neural architectures for semi-structured data, learned components for document databases, and generative models that work natively with hierarchical data — the research substrate that underpins our agent systems.
How We Work
We practice what we research. Our lab runs on automated pipelines that triage issues, route tasks to AI agents, and track experiments with minimal human intervention. The same infrastructure we build as research tools powers the lab itself.
Every process that can be delegated to an agent is. Humans drive prioritization, design decisions, and interpretation — the mechanical work of running experiments, reviewing code, and managing workflows is handled by the systems we build. This isn't a future aspiration; it's how we operate today.
Latest Posts
Breaking Through Tabular Constraints for Synthetic Data Generation
Most synthetic data generation tools assume flat tables. Real-world application data is often nested JSON with optional fields and variable-length arrays. The ORiGAMi architecture is the first to handle semi-structured data natively.
Loosening the Reins
AI is moving so fast that we're becoming the bottleneck. On LLMs as idea search engines, the reversal of control, and where this all goes.