Operating model
Define outcomes, ownership, team structure, decision rights, measures, and governance for employees and agents working as one system.
AI operating models and agentic systems
Building AI-first organisations
We redesign your highest-value workflows for people and AI agents working together, then build the operating model and software that let your organisation grow faster than its headcount.
Your first engagementA pilot delivered in 2–3 months
One core workflow, such as order handling or billing, redesigned end to end
Purpose-built software that replaces costly licences or a legacy in-house system
A clear plan for every step: which tasks agents complete autonomously and which decisions stay with your employees
Your team trained to operate the new workflow
The first building block of your new AI operating model
For leaders replacing core platforms, redesigning operations, or moving AI from experiments into production.
AI does not transform a business when it is bolted onto existing processes.
Most organisations add copilots and point solutions to their existing processes. A tool that reads incoming orders removes the data entry, but people still check, price, book, and correct each order downstream. The cost per order barely changes.
We do not automate your existing process. We start from the outcome you require, such as booking an order or issuing an invoice, and redesign the process so agents handle the routine steps and your team takes the decisions that call for human judgement.
We redesign your processes and build the software.
Three connected capabilities that take one valuable workflow from redesign to production.
Workflows and operations
Redesign one workflow end to end. Remove avoidable handovers, assign each task deliberately, and design human review around risk and exceptions.
AI-first products and platforms
Build or modernise the software, data foundations, and agent workflows that run the new operating model. You choose which AI model handles each task and where your data is processed.
Illustrative example
From a manual task to an accountable AI-first workflow
Booking an incoming order shows how routine work changes and where people remain accountable.
Before
An order arrives by email or as a PDF. A person in the sales office reads it, looks up each article, types the order into the system, checks stock and the customer’s credit limit, and releases it.
AI-first
The incoming email starts the workflow. An agent reads the order in whatever format it arrives, matches the customer and each article, and books it within the agreed credit and stock rules. An unknown article, a price deviation, or a breached credit limit is routed to a person. Every step is logged for audit.
Book an incoming orderIllustrative
People
Agents
System
Policy checkAudit trailReceive orderRead orderMatch articlesBook orderReviewexceptionLog for audit
Routine orders are booked without manual re-entry. Your sales office sees only the orders that require a decision.
We prove the change on one workflow before you scale it.
A first engagement runs for two to three months, from the first interviews to a running pilot. That is comparable to a typical strategy project, but it ends in working software.
- 01
Diagnose
We agree the goals and constraints of the project with you, such as which data may leave your organisation and which decisions require human judgement. We interview the employees who run the process every day and analyse the systems and data involved. We then select the first workflow together.
Output A shortlist of workflows ranked by value and feasibility, today’s cost and speed as a baseline, and a recommended first workflow, including whether to keep, buy, or build each system it touches
- 02
Redesign
Map the outcome end to end and break it into tasks. Assign human, agent, and system responsibility, including controls and exceptions.
Output A target workflow, operating model, and build plan
- 03
Prove
Our forward-deployed engineers build the pilot inside your organisation, alongside the team that will use it, then measure speed, cost, quality, and how often your employees have to intervene.
Output Measured results from your own processes and a business case for the decision to scale
- 04
Scale and transfer
Extend the proven workflow to additional sites, teams, or processes, establish reusable foundations, and equip internal teams to operate and improve it.
Output Lasting capability rather than permanent supplier dependence
One team from operating model to working system
Consultancies hand over a strategy. Software vendors sell their version of the process, and agencies build the one you already have. We redesign the process, build what is specific to you, and use standard products where they fit.
Strategy and delivery together
One partner leads your project from the first conversation through to production. The team that redesigns your processes also builds the software, so nothing is lost in a handover between consultants and developers.
Outcome before technology
We agree your performance targets, such as cost per case, processing time, or error rate, and measure against them. We earn nothing from licences, and the diagnosis decides whether to keep, buy, or build each system before we propose any build.
Small senior teams
An employed core team does the work. Specialists join for peaks, so capacity follows the project instead of becoming permanent headcount.
Forward-deployed engineers
Our engineers work with the employees who run the process and take the software into production inside your organisation. They use AI across the development lifecycle to increase speed and reduce delivery overhead.
Built for transfer
Before the build starts, we agree how your team will run and maintain the platform, and what support you require from us. Architecture and documentation are designed for that handover.
Control by design
Permissions, review thresholds, observability, and audit trails sit inside the workflow. Each agent task is tested against real, solved cases before it goes live and before any AI model is changed.
Built for consequential change
FORM II is for organisations with a valuable workflow, core platform, or operating constraint that cannot be solved by adding another tool.
For private equity owners, see what an owned, AI-first core platform does for a portfolio company.
Signals of a strong fit
- A senior sponsor who can decide on both the process and the IT systems
- High licence fees, or low-performance legacy systems
- Access to users and domain experts
- Readiness to change process and technology together
- A need to move before internal capability can be built
- Critical know-how held by a few experienced people
From the blog
All postsStandard software is someone else’s operating model
Software shapes how organisations work. Now, AI is rapidly changing both what it costs to build software and what it can do, which reopens a question most mid-sized organisations considered settled: should you buy or build your core IT systems?
Which process would you redesign first?
Tell us which workflow or system you want to reimagine and what it costs you today. We will get back to you within two business days, and if we can help, we will arrange a call to explore options together.