Private AI you own.
Private AI runs inside your own walls. Your data and your prompts never leave your network, are never seen by an outside cloud, and never train someone else's model. And you own it.
What it is
Private AI, built for regulated work.
Private AI means running AI in a closed environment, on dedicated hardware inside your own network, so your data and prompts never leave your control and no outside platform can see them or train on them. Built for the sector, with a person keeping the judgment.
Private by design
The AI runs on dedicated hardware inside your own environment, never a shared public cloud. Nothing is handed to a third party.
You own it
The delivered system, your data and your intellectual property stay yours, inside your walls. You are not renting someone else's model.
A price you can budget
A scoped build, a fixed monthly fee, and no usage fees. Predictable however much your people use it.
Consultancy-led
Discovery and design come before any build, so every deployment reflects how your people actually work, not a generic template.
For the finance function
Close faster. Report sooner. Scale without hiring.
A team of specialised finance agents working alongside your finance function, built by people who have run finance, and deployed inside your own environment.
Reporting
Board packs, management accounts and variance commentary, drafted from live system data, in your format, every cycle.
Invoicing
Generated, validated and dispatched, from purchase order to revenue recognition, with no manual touchpoints.
Payroll
Calculated, compliance-checked and processed every cycle, with the statutory rules handled by default.
Accounts receivable
Chasing, payment matching, dispute triage and debtor reporting, automated end to end.
Accounts payable
Invoice capture, three-way matching, approval routing and payment. Your team reviews the exceptions only.
Procurement
Purchase orders, vendor management and spend control, run against policy, with a full audit trail.
A layer above your ERP. Connected, not converted. Nothing migrated, nothing replaced.
Private by architecture. The agents run inside your environment, and your ledger never leaves it.
The finance team keeps the judgment. The agents absorb the repetitive work; people review, approve and decide.
The IP stays yours. The models, rules and learnings built on your data belong to you.
It begins with a short, fixed-scope diagnostic: your processes and systems mapped, the highest-value agents identified, and a build plan you can take or leave. The first agent goes live in weeks, and the platform grows one agent at a time.
For schools
The school that owns its AI.
The same private AI platform, proven in another regulated sector. Schools need AI, but consumer tools cannot be trusted with student data, do not know the curriculum, and forget every student the moment a session ends. SchoolAI is the opposite: a private platform the school owns and controls.
AI Co-Teacher
Trained on the school's own curriculum. Guides students by questioning rather than giving answers away, and supports lesson planning and differentiation.
AI Tutor
Personalised revision, exam preparation and adaptive assessment, aligned to the school's syllabus and mark schemes, adjusting difficulty in real time.
AI Student Wellbeing
Early signals on pastoral concerns, wellbeing check-ins and confidential support pathways, with safeguarding designed in, not bolted on.
Role-based portals. Students, teachers, parents and administrators each see exactly what they need.
It remembers each student's journey. Progress, gaps and mastery across terms and years, not just one session.
Teacher content stays in an encrypted vault. It is never used to train an outside AI.
Adaptive assessment. It checks understanding before a student moves on.
It connects to what the school already runs. The existing learning-management and reporting systems.
Built for GDPR, APP and POPIA. Student data never leaves the school's environment.
The first premium deployment is under way at an independent college in Melbourne, Australia.
Why now
Most AI lives on someone else's cloud.
The cost follows usage
AI on the public cloud is getting more expensive and harder to budget. The bill follows how much you use it, not how many people you employ.
The rules are tightening
Data-protection rules are tightening on where sensitive information may be processed. For finance, putting client data on a shared cloud is getting harder to justify.
AI is coming in-house
Organisations are bringing AI back inside their own walls. That is the shift TransformationWorks was built for.
Who is behind it
Run like a business, built in-house.
Private AI is only worth owning if the people behind it are real. Here they are.
Kenneth Tan
Technical co-founder · Silicon Valley & Singapore
Designs and leads the build of every deployment, through his Singapore firm MyLearnZone. More than ten years of private AI deployment, built in-house, not outsourced.
mylearnzone.com ↗Simon Gallagher
Founder, TransformationWorks · Chartered Accountant
Built TransformationWorks from the ground up. A chartered accountant with thirteen years in international C-suites, including Regional CFO of Adecco ANZ. An AI company that speaks the language of the boardroom, not the server room.
transformationworksgroup.com ↗Fransua van der Westhuizen
Kanton-West · Exclusive South African licence
Holds TransformationWorks' exclusive South African licence. Scopes each engagement, appoints the specialist partners who deliver it, and stands behind the outcome here.
LinkedIn ↗MyLearnZone, trusted by
The South African licence
Scoped here. Built globally.
Kanton-West holds, through SATalent1st, the exclusive South African licence for TransformationWorks. One appointment, covering the whole market. For a South African business that means Private AI with a local principal: scoped and stood behind here, while the engineering, across Silicon Valley, Singapore and Melbourne, builds and maintains it.
The private AI is a capability brought to the table. The finance work itself is still done personally.
This site is Kanton-West's own work, built in-house with AI in the loop, from concept to live deployment.
Private AI, on your terms.
A scoped build, a fixed monthly fee, and no usage fees, inside your own walls. Outline the situation and what you need.