Company.
Our engineers use AI and one model of the plant to deliver the engineering work manufacturers already buy. Each project adds to the model, so the next one starts with more of the site already known.
AI makes engineering faster. On an existing plant, speed without evidence is how changes go wrong.
That is why every answer we give is checked against your plant’s own evidence and the physics around the change before it reaches your site. Where the evidence runs out, we say so.
Our forward deployed engineers do that work. They embed in your organisation, walk your site, sit with your project and maintenance teams, and take each question to an answer your engineer of record can sign.
SyntaGraph™ is how they work. What they verify stays in one model of your plant, so the next decision starts with more of the plant already checked.
What we sell.
A manufacturer asks whether a machine fits, whether there is utility capacity, whether a project is ready to tender, whether the as-built drawings are true or whether the site can take a transfer.
We deliver the answer and the evidence behind it. There are no seats or licences to buy. You keep your systems of record, and the site model exports in an open format.
SyntaGraph™ on our synthetic reference plant.
What comes next.
AI is taking on more of the extraction, reconciliation, analysis cases and drafting. Our engineers keep responsibility for professional judgment.
Today
Understand and verify
We deliver feasibility, existing conditions, handover, change, turnaround and due diligence studies on the plant as built.
Next
Design on the verified plant
We will design from a reconciled plant, with our own engineer of record.
Later
Evidence across the plant's life
We will add verification, qualification and inspection on the same record, each one once we have the technical and regulatory capability for it.
How we work.
- 01
Understand before predicting
- 02
Predict before recommending
- 03
Simulate before executing
- 04
Execute only within governed boundaries
- 05
Learn from every outcome
Accountability and data.
Accountability
An engineer signs
AI extracts, reconciles and drafts. Where professional judgment or statutory accountability is required, a qualified engineer stays responsible.
Data
Inside your boundary
Your data trains nothing without a permission you negotiated. Any reuse needs a separate written agreement.
Gaps in the evidence
VerifiedNot Demonstrated
Findings the evidence supports are marked Verified. When the evidence cannot carry a conclusion, the deliverable says so and names the evidence that would close the gap.
The name.
In our name, normality means the latest verified and permitted state of a site's systems, spaces, assets and operating conditions. What was designed is not always what was built, and what was built is not always how it runs. We show you those differences before you commit capital.
Where we work.
Pharmaceutical manufacturing
Feasibility, baselines and change assessments for GMP manufacturing sites.
Data centres
Feasibility, baselines and change assessments for data halls.
Coming soonBecome a design partnerFood and beverage
Feasibility, baselines and change assessments for food and beverage plants.
Coming soonBecome a design partnerWe are looking for design partners in data centres and in food and beverage. If you have a change coming up on one of these sites, help us shape the service around it.
Become a design partnerStart a study.
Tell us the site, the project stage and the decision you need to make. We will propose the deliverable, the scope and the evidence it needs.