Research
Architecting cognition
Our research focuses on the systems-level question: how do you build AI that learns continuously, reasons reliably, and augments human capability in complex domains?
Principles
How we think about research
Architecture over scale
Intelligence emerges from how components are structured, not just how large they are. We invest in the right abstractions.
Deployment as research
Deployment constraints that don't appear in benchmarks inform our research. Every engagement advances the systems.
Measure what matters
We optimise for R&D throughput, reasoning depth, and consistency under complexity -- not benchmark scores.
Focus areas
Memory, reasoning, learning, verification
The core components we believe are necessary for systems that genuinely augment human intelligence.
Hierarchical memory architectures that mirror human cognition -- working memory for immediate context, episodic memory for experiences, semantic memory for consolidated knowledge.
Publications
LEAP: Learning through Episodic Accumulation and Processing
Surprise-Gated Memory Consolidation in Neural Networks
Multi-step reasoning through hypothesis generation, refinement, and verification. DAG-based execution with full trace logging for interpretability.
Publications
Monad: A Traceable Runtime for Agent Orchestration
Iterative Refinement in Multi-Document Reasoning
Systems that improve from deployment feedback without catastrophic forgetting. Meta-gating mechanisms that balance stability and plasticity.
Publications
Engram-VQ: Fast/Slow Memory with Meta-Gating
Diffusion-Based Reasoning at Small Scale
Deterministic replay, audit trails, evidence chains, and failure clustering. Systems that can explain their reasoning, reproduce their outputs, and trace decisions back to source evidence.
Publications
Provenance: Deterministic Decision Replay for Agent Systems
Failure Clustering in Multi-Step Reasoning Pipelines
Access
Research artefacts
Access to our research artefacts, tools, and early-stage systems is available to partners and collaborators. Request access to explore integration with your workflows.
Request AccessAvailable with access:
Collaborate with us
We're always looking for collaborators -- whether you're a researcher, practitioner, or organization working on related problems.