GenAI guidance in force
The OAIC has published guidance on privacy and the development or training of generative AI models. The deployment's data path is written for that review.
Self-hosted AI / Australia
Self-hosted AI for Australian companies, where the AI review begins with OAIC guidance.
Private AI deployed for Australian businesses under the Privacy Act 1988 and the Australian Privacy Principles, with hosting in Sydney and every model provider named. The OAIC has published guidance specifically on generative AI and privacy, and the deployment is designed for that review.
From AUD 5,500 / $3,500 / NGN 2.5m floorsap-southeast-2 (Sydney) optionClient-owned accountsFull handover
Example only / architecture varies by client
Designed for review against the Privacy Act 1988 and the Office of the Australian Information Commissioner (OAIC)ap-southeast-2 (Sydney) optionClient-owned billingFull handover
Why it matters here
Australian procurement reviews AI as a data-flow decision.
OAIC guidance, sector obligations, and state law can all apply to the same workflow. The deployment documents one path for all of them.
Australian Privacy Principles
APP entities must be able to show how personal information is collected, used, disclosed, and secured. Named processors and documented access support that.
Data in Australia
ap-southeast-2 (Sydney) is the default region, with fully local deployment for stricter requirements.
The three levels
The architecture follows the constraint.
There is no single correct answer. Quality, privacy, residency, and budget decide the path.
Your cloud, in Sydney
App and document store in your account in ap-southeast-2 (Sydney), with the model on your key. No GPU required.
For location requirements
An approved cloud region or managed service when residency and procurement matter.
For strict control
Model, embeddings, and retrieval on your hardware when external inference is not acceptable.
What it really takes
Measured workloads, not GPU shopping lists.
Hardware is sized after the workload is real. These are starting points, not quotes.
4 to 8 vCPU, 16 to 32 GB RAM
No GPU required for the app and document store. Model usage bills stay on the client account.
From 24 GB VRAM
A starting point for a small quantized text model. Model size, concurrency, and latency decide the real requirement.
Power, cooling, updates, backups
A GPU is one part of the environment. A partner data centre can be more realistic than an office server room.
Not the first promise
We size hardware after the workload is proven. No GPU inventory before a signed scope and deposit.
Straight answers
Common questions.
Where would the data be hosted?
The default is ap-southeast-2 (Sydney). The architecture is documented in advance if another region or fully local deployment is required.
Does the OAIC approve systems?
No. The OAIC does not approve products. The deployment is designed to be reviewable under the Privacy Act, the APPs, and OAIC's generative AI guidance.
Who owns the accounts?
The Australian client owns the cloud, model, and storage accounts. Pristine3D hands over credentials, the runbook, and the training session.
Do invoices come in Australian dollars?
Yes, from AUD 5,500 for a first workflow, with USD and naira floors unchanged.
Australian AI procurement is a data-flow review. The deployment is written for it.
Why trust Pristine3D?
We build and operate production software.
Pristine3D Ltd builds and operates live digital products, and we run private AI workflows internally as part of our own operations. We scope around your real workflow: the documents you own, the questions your team asks, and the access boundary you approve. Based in Lagos, Nigeria, we work remotely with clients worldwide.
We already run live products
Based in Lagos, Nigeria, Pristine3D currently builds and operates smartcards.ng, venu.ng, photoshoot.ng, and ugc.ng in production.
We start with the actual workflow
One input, one output, one test set, and one person who owns the result. We scope a real workflow instead of a transformation programme.
The boundary stays visible
Cloud, model, storage, and messaging accounts stay in your name. The chosen data path, access rules, test record, documentation, and training are part of the agreed scope.
Pricing / fixed scope
Know the starting numbers before you ask.
The final quote follows the workflow. Infrastructure and model bills stay on your accounts.
One Workflow
Starting from$3,500 / ₦2.5m
2 to 3 weeksOne workflow for a small team, with training and handover.
See the one-workflow packageFirm-wide
Starting from$7,500 / ₦8m
4 to 6 weeksMultiple sources, roles, integrations, and admin handover.
See the deploymentAnnual support
Starting from$3,000 / ₦1.5m
per yearStandard care for one delivered workflow. Optional. Larger deployments and active monitoring are separately scoped.
See supportArchitecture review from $500. Standard annual support is $3,000 / ₦1.5m per year for one delivered workflow. New workflows, integrations, active monitoring, and infrastructure are separately scoped; infrastructure, model, storage, and messaging bills stay on client accounts. Full pricing and what changes the quote
Own your knowledge base
The model is not the product. The knowledge base is.
Documents, the retrieval index, access rules, and the workflows built around them are the asset, and they compound. We deploy so the knowledge base stays yours: on your accounts, in the environment you choose, under access rules your team defines. The model behind the answers is a connector, so the knowledge base moves with you, not with a vendor.
Your corpus, your index
The document store, metadata, and retrieval setup live on accounts you own. No vendor holds the corpus.
Models are swappable parts
Change the model provider, move regions, or go local without rebuilding the knowledge base or the workflow.
The knowledge base is the alpha
Every improvement to the corpus improves the answers, and the improvement stays with you, not with a vendor.
Keep exploring
Related setups.
Start with the constraint
What is your constraint?
Residency, OAIC review, procurement, or a no-external-inference rule. Describe it and we will map the architecture.