Work we do
alongside
the products.
Facilitation, software development and AI systems work. Most engagements start because a partner needed something the roster did not cover, and several products began life as exactly that kind of engagement.
Getting a room to build something.
Run on Strategy Spaces, so the phase track, timing and expert routing are the same tooling we ship. Delivered in person, online, or both.
Innovation session facilitation →
Half-day to multi-day sessions where a group works a real problem rather than discussing one. We design the sequence, run the room, and leave you with the artefacts and the decisions written down.
- Half day to 3 days
- 8–80 participants
- In person or online
Sprint build sessions →
A team ships a working prototype inside a week. Framing on Monday, something people can actually use by Friday, with our engineers building alongside yours rather than for them.
- 3–5 days
- Working prototype
- Your team included
Useful, governed, and accountable.
The same constraints we hold in the products: models suggest, people decide, and every output is traceable to a source.
AI orchestration and prototyping →
Multi-step and multi-model systems built to do a specific job, with the retrieval, tool use and human checkpoints designed in. We prototype quickly, then show you what it costs to run properly.
- Agent orchestration
- RAG pipelines
- Evaluation harness
- Cost modelling
AI productivity, governance and ethics →
Practical adoption work: where AI genuinely helps, where it should not be used, and what has to be true before it touches anything sensitive. Includes staff capability building and a written governance position.
- Adoption roadmap
- Governance framework
- Staff capability
- Risk assessment
AIOps, ModelOps and ethical systems design →
Getting models into production and keeping them honest: deployment, monitoring, drift detection, retraining triggers, and audit trails that satisfy a board as well as an engineer.
- Deployment
- Drift monitoring
- Audit trails
- Retraining
Software development →
Full delivery for research, education and public-purpose software. Local-first where it matters, open formats throughout, and documentation written for the people who will maintain it after us.
- Web platforms
- Data tooling
- Open formats
- Handover included
The gaps in a dataset are usually the finding. Mapping work should make them visible rather than smooth them over.
Mapping and GIS solutions.
Built on HyveMaps or standalone, depending on what the work needs.
Mapping and GIS solutions
Coverage, gap and overlap analysis on real public data. Custom layers, field collection, and provenance recorded on every dataset so a map can be defended in a submission or a hearing. Consent terms and uncertainty are carried through to the published view rather than dropped at export.
- Coverage & gap analysis
- Custom layers
- Field collection
- Provenance & consent
- Open data publishing
Deployment and capability work.
Three further engagement types, each with its own page.
Robotics deployment →
Getting robotic systems from a working demonstration into a real site, with the safety documentation and staff training that makes it stick.
- Site survey
- Fleet operations
- Safety documentation
- Staff training
Intelligent automation →
Automating genuine drudgery without hollowing out the judgement work around it. Process mapped first, humans kept in the loop by design.
- Process mapping
- Document handling
- Triage and routing
- Measured savings
R&D consulting →
Research and development where we work inside your team under NDA. Nothing appears in our case studies and the intellectual property is yours.
- NDA as standard
- Embedded
- Your IP
- Documented handover
Innovation augmentation →
Raising what an existing innovation team can do, rather than doing it for them. Structured so the capability stays once we leave.
- Embedded coaching
- Method and tooling
- Planned taper
Who pays, and who doesn’t.
Commercial engagements fund the discounted and gratis work. That is the whole model.