Lily LabsContact →

What we do for clients

Seven engagement types. Each is informed directly by the research programmes, and each is shaped to the constraints of the audiences we work with.

AI strategy consulting

Architectural review for organisations choosing between black-box statistical AI and explainable, structured alternatives. We assess fit, risk, regulatory exposure, and total cost of ownership — including the costs that surface only after deployment.

For
CTO / CIO / Head of AI
Outcome
Decision-grade architecture recommendation, defensible to your board and your regulator.

Explainable AI system design

End-to-end design for AI systems that must justify every decision they make. Suitable for financial services, healthcare, legal, insurance, and any environment where audit is non-negotiable.

For
Regulated-industry product and risk teams
Outcome
System design with auditable reasoning paths and compliance encoded as first-class structure.

Knowledge graph architecture

Design and implementation of structured, queryable, governed knowledge stores to replace or augment opaque LLMs. We bring patterns from our geometric reasoning research and from decades of enterprise knowledge work.

For
Engineering teams replacing or de-risking LLM-only stacks
Outcome
Production-ready graph architecture that you own, can audit, and can run on hardware you already have.

Critical national infrastructure AI assessment

Security and resilience analysis of AI applied in industrial control systems, SCADA, and operational technology environments. Includes air-gap-compatible deployment review and standards-database integration.

For
CNI operators · government CNI assurance
Outcome
Written assessment of AI-related risk in your OT estate, with prioritised remediation.

GEO and AI search readiness

Generative Engine Optimisation — making your organisation visible to AI search engines (ChatGPT, Perplexity, Claude, Google AI Overviews). Delivered through our consumer arm, R U AI Ready.

For
Marketing and digital teams
Outcome
AI-readiness audit and prioritised implementation plan.
Visit

Bespoke research engagements

First-principles work on hard problems. We take on a small number of bespoke research engagements per year — typically multi-month, output-defined, and with the client's name removed from the public-facing research where required.

For
Government · enterprise · investors
Outcome
Original research output, IP arrangements negotiated per engagement.

AI-collaborator methodology consulting

Methodology design and deployment review for organisations using AI collaborators on long-horizon technical work — derivational science, engineering analysis, or structured research programmes where evidential discipline matters across months of work. Based on our own six-month in-production pattern, documented across ~100 sessions.

For
Research programme leads · R&D heads · AI governance
Outcome
Working pattern documented for your domain, with audit-before-commit and tool-grounded recall baked in — designed to hold up under peer review or regulatory scrutiny.
Discuss this

Engagement model

We take on a small number of engagements at any given time. Most begin with a fixed-scope discovery — typically two to four weeks — after which the shape of the longer engagement (if any) is clear to both sides.

Lily Labs is set up for serious work, not for procurement theatre. Initial enquiries are handled by Steve directly.

Discuss a project