Getting started
Start with one recurring failure, not a platform rollout.
A pilot is not a trial licence and a login. It is a scoped piece of real work with criteria agreed up front, so that at the end there is something to decide rather than an impression to argue about.
Ingestion filters
Set by your admin
By file name
Files whose name matches are skipped: never read, never indexed.
By content
A document has to be read to be classified. A match is discarded before indexing, so it never enters operational memory and can never be surfaced in an answer.
Financial
Budgets, margins, invoices and anything with commercial figures.
HR and personal data
Payroll, personnel files and documents with identifiable personal data.
Legal and contracts
Contracts, NDAs and legal documentation.
Project Auroracustom
Written in plain language: skip anything covering the Aurora commercial agreement.
You draw the line on what may enter the knowledge base.
How a pilot works
Narrow on purpose. Measured from day one.
The value of this product comes from density of real activity, not from breadth of coverage. One line used every day produces more evidence than four sites and every asset.
Determine scope and priorities
One site — the plant where the pain is loudest. One recurring failure, or one family of equipment the plant runs several of. One workflow, usually incidents and shift handover. We agree the success criteria and the baseline before anything is installed.
Build and prepare
About a week, most of it ours. The installed base is imported from your ERP or CMMS, limited to the scope. Your documents go in as they are and Norlan proposes which models each belongs to. Ingestion filters are agreed first, so nothing is read that should not be.
Launch and run
Real shifts, real incidents, on the phone at the machine. We are in the thread with your team the first weeks — not to document for them, but to make sure asking the machine becomes the reflex before phoning the senior.
Measure and decide
The numbers you agreed, measured against the baseline — MTTR on the failure in scope among them. If technicians did not use it during real work, the pilot has failed and we will say so. Expansion is a separate decision, taken after the review, not a renewal that happens by default.
What gets put in place
A week of preparation, most of it ours.
What we need from you is an export and a decision about what may be read.
- Installed base imported, keyed by asset code, so a re-import never duplicates
- Documentation loaded as it is — manuals, procedures, schematics, however the folders ended up
- Relevant history brought across, so the first questions have something to stand on
- Ingestion filters agreed before anything is indexed — payroll, commercial, legal, anything you name
- Preventive rounds imported from the CMMS, so the first line workflows run on real work
What gets measured
Agreed before we start. Against your own numbers.
These are questions to agree, not results to promise: we have none to publish yet, and saying otherwise here would be lying exactly where it matters.
- Time to understand an open incident after a shift change
- Share of closures that carry a structured root cause and a verified resolution
- Later incidents that reuse an earlier resolution
- Recommendations accepted versus rejected by technicians
- Interruptions, calls or messages to a senior technician avoided
- Time to find the relevant manual passage or prior intervention
- Assets or models that depend on a single expert, before and after
- Knowledge promoted from an intervention into an approved guide
- Similar-issue suggestions shown, and how many the technician opened
What your technicians see on day one
One hour at the machine is the onboarding.
No three-day training, no content workshop before anyone gets value. The thread, the assistant and voice are the whole surface on the first day — and the manuals are already in it.
The machine's thread
Open the asset, see what happened on it, report what is happening now. The group chat, with a memory.
The assistant, with the manuals already loaded
Ask the machine. It answers from the manual you gave us last week, cites the page, and says when it is answering from the trade instead.
Voice
Talk, take a photo, keep fixing. Norlan transcribes and keeps it on the asset. Typing is optional.
When you have questions
You talk to the people who build it.
The people who build it
There is no tier-one support desk between you and the product. During a pilot you talk to the team that builds Norlan, in the thread, by phone or by email.
Your CMMS, untouched
Nothing is migrated or switched off. Work orders, planning and traceability stay where they are; Norlan holds the reasoning your CMMS was never designed for.
Written answers
Any question about data, security or what is not in place yet gets answered in writing, before production data flows.
Questions about starting
How long does a pilot take?
What do you need from us to start?
Do our technicians need to create content before it is useful?
What happens to our data if we stop after the pilot?
Do we have to change how technicians work?
Book a demo
A 30-minute walkthrough built around the plant's machines, not a canned demo.
We start with the operation — which failures repeat, who gets called when something breaks, what the CMMS does not hold — and then show how Norlan would work on it.
What we usually cover
- How an incident flows today, from detection to fix
- Where the conversation actually happens: the systems, or calls and WhatsApp
- What takes longer — the diagnosis, or getting the right people and information together
- How much of what a technician knows is written down, and what happens when they are away
- Which tools are really in play: CMMS, ERP, mobile, paper
- What a pilot would measure, and how
Pick a time that suits the plant. Bring one failure that keeps coming back.
The call is with the people who build the product, with no qualification call first. Norlan is pre-general-availability and shaped with maintenance teams in pharma and food processing.