Tech
The machine provides, the human decides.
AI engineering, AI ERP and governance AI on your own data, with a person signing off on every decision that matters.
AI engineering
Agents, retrieval and automation on your data and systems, tested and running in production.
We find the work in your company that AI can take over and build it: agents for routine tasks, search across your documents and data, and automations between the tools you already use. The agents prepare the facts and the next step; your people approve it. Everything is tested on your real data before it goes live.
AI ERP
ERP workflows with AI built in, connected to the systems you already use.
We design and build ERP workflows for fashion and trading companies, from purchase orders and production tracking to shipping and invoicing. AI reads the documents, fills in the records and flags what needs a decision, and your team makes it.
Governance AI
Model inventory, controls and an audit trail, mapped to the NIST AI RMF, ISO/IEC 42001 and the EU AI Act.
We set up the inventory of your AI systems, the controls around them and the records that show they work, so you can answer customers, auditors and regulators. The audit trail shows what each system proposed and who signed it off.
UI/UX design
Interfaces and products that people understand the first time they use them.
We design the flows, screens and design systems for your tools and products, and test them with the people who use them.
What we advise on
Which work to hand to AI first, what to build and what to buy, which decisions stay with your people, and how to run AI safely on your own data and systems.
What we deliver
- Working software, running in production
- Agents and automations, with sign-off steps for your people
- Integrations with your ERP, email and documents
- Governance records: inventory, controls and audit trail
- Interface designs and design systems
- Documentation, training and handover
How a project runs
- A first call about the process and the result you need.
- A short study of your data and systems, and the result agreed up front.
- A working prototype on your real data.
- Build, tests and rollout, with your people approving each step.
- Handover, training and the result measured in your numbers.
Every decision leaves a thread you can follow.
Pick a decision and follow it from the data to the person who signs it off. Where a rule stops it, nothing changes until a person decides.
- Data
- Competitor prices, this week
- Model
- Pricing model v1.8
- Rule
- Discount limit exceeded, stopped here
- Person
- Sent to the CFO for a decision
Result: stopped at the rule. Nothing changed without a person.
Illustrative examples, not client work.