Services
Five ways into one capability system.
Services are not products on a shelf. Each one is a door into the same operation: your workflows, your systems, your data, and the people who keep authority over decisions.
The five services
01. AI adoption & opportunity
Most businesses know AI matters. Far fewer know which workflow to start with, what it would cost, and what should stay human.
An opportunity assessment is a structured look at how work actually moves through your business: the volumes, the hand-offs, the delays, the rework and the decisions that require judgement. We map that against what current AI, automation and integration can reliably do, and against what your systems and data can support.
See the full service →many signals · few candidates · one to prove first
You leave with
- An honest map of where AI fits, and where it does not
- One opportunity chosen to prove first, with a defined result
- A sequence the business can fund step by step
Signals this is the door
- A board or leadership request to "do something with AI" without a defined use case
- Several small pilots running, none connected to a production workflow
- Vendor proposals that are hard to compare because the underlying problem is unclear
02. AI agents & copilots
An agent is only useful when it is given a real job, real access and clear limits. We build agents that way.
Every agent we build starts with a written description of the role it supports, the tasks it takes on, the sources it may read, the actions it may take and the point at which it must hand back to a person. Only then do we choose models, retrieval methods and integrations.
See the full service →ROLEOrder-desk copilotKNOWSprice list · stock · termsCANdraft · check · prepareCANNOTsend · discount · commit stocka job description before it is software
You leave with
- A working agent with a written boundary of authority
- Connection to the knowledge and systems it needs
- A review loop so it improves without drifting
Signals this is the door
- Experienced staff doing repetitive research, drafting or summarising before real work starts
- Internal questions that are answered by asking a colleague rather than a system
- Institutional knowledge locked in documents, tickets, email threads and past proposals
03. Workflow automation
The fastest capability gain in most businesses is removing the work that should never have needed a person.
We map the workflow end to end, including the informal steps that never appear in the official procedure. Then we decide, step by step, what should be deterministic, what needs AI to read or classify unstructured input, and what must remain a human decision. Most production automations are hybrids, and the most reliable parts are usually the simplest.
See the full service →■ automated step · ● exception, routed to a person
You leave with
- Workflows that run without being chased
- Exception rules written by your team, not guessed
- A record of every step, every time
Signals this is the door
- The same data entered into two or more systems by hand
- Approvals that depend on someone remembering to chase
- Documents moving through inboxes, shared drives and messaging apps
04. AI & system integrations
An AI system that cannot read from your records or write back to them is a demonstration. Integration is what makes it a business system.
Integration work is unglamorous and decisive. We start by understanding each system's data model, API capabilities, permissions and rate limits, then design the data flows so that AI components receive accurate context and their outputs are written back with the right validation. Security, data residency and access control are decided at this stage, not retrofitted.
See the full service →one version of each fact, across systems
You leave with
- One version of each fact across systems
- Fewer re-keyed fields, fewer reconciliation days
- A foundation agents and automation can trust
Signals this is the door
- AI pilots that stall because they cannot access live business data
- Staff pasting information between systems and AI tools by hand
- CRM and ERP that disagree about the same customer or order
05. Intelligent business systems
When a workflow needs more than an agent or an automation, it needs a system designed around it, with people, controls and visibility built in.
An intelligent business system is designed from the workflow outward. We define the states work passes through, the decisions taken at each state, who takes them and what information they need. AI is placed where it interprets, drafts or recommends; deterministic logic handles what is well defined; people handle what requires accountability. The result is one place from which the operation is run and observed.
See the full service →records · rules · roles · reporting
You leave with
- A system shaped by the operation, not the other way round
- AI built in where it is useful, not bolted on
- Something your team owns and can grow
Signals this is the door
- An operational process managed through spreadsheets and messaging because no system fits it
- Managers who cannot see workload, status or exceptions without manual reporting
- Several automations and AI tools that have grown up separately and now need coordination
In what order
A sequence that follows the work.
Most engagements begin with a question about where AI could help. The answer usually involves more than one service, in an order set by the workflow rather than the technology.
01
Find the opportunity
AI adoption & opportunity
02
Connect the systems
AI & system integrations
03
Automate the routine
Workflow automation
04
Add judgement where inputs vary
AI agents & copilots
05
Run it as a system
Intelligent business systems