An AI consultant for Australian teams, Sydney to Melbourne.
Sentry AI builds AI into the real work of Australian businesses, with inference kept in Australia where the data needs it and the Privacy Act scoped in from the first conversation. We are based in Auckland, in person in Sydney and Melbourne by arrangement, and train your team in your own office.
The short answer
Sentry AI works as an AI consultant for Australian businesses from Auckland: we deploy Claude with every inference kept in Australia where the data requires it, scope the Privacy Act and the Australian Privacy Principles in before anything is built, train the team in their own office in Sydney or Melbourne by arrangement, and stay on remotely as their AI partner.
Every answer, skill and training day comes from James Oldham, who founded Sentry AI in Auckland and does the engineering on the work these pages describe. About James
- people connected to the AI systems we run
- 185+
- skill runs a month across client teams
- 1,500+
- skills in a single client's library
- 45
- organisations we run AI for today
- 7
Live figures from the teams we work with, October 2026.
Case study: data kept in Australia
Caregroup: Claude across 15 practices, every inference in Australia
A healthcare network of 15 practices and 180,000 patients needed frontier AI in front of clinical staff without patient data leaving the region. Claude on every seat points at the network's own gateway on AWS, which sends every inference to Amazon Bedrock in Australia. Nothing goes to a vendor's own cloud.
Claude with inference kept in Australia
The first question an Australian team asks about AI is usually where the data goes. For most work, the answer can be simple: run the model in an Australian region, on infrastructure the business controls, under the same access controls and audit trail as the rest of its systems.
That is how the Caregroup deployment works. Claude Desktop on every seat talks to the network's own gateway, and the gateway sends every request to Amazon Bedrock in Australia. The team gets the same Claude, and the board gets an answer to the residency question it can check. Where the data is not sensitive, a standard Claude plan is often the simpler and cheaper choice, and we will say so.
What an AI consultant should scope in for Australia
None of this is a reason to wait. It is a reason to settle a few questions before the first workflow is built rather than after it is in use. We are not lawyers, and where the answer turns on legal advice we will say so and bring yours in.
The Privacy Act 1988 and the APPs
The Australian Privacy Principles apply to personal information whether a person or a model handles it. Before a workflow touches customer, patient or staff records, we settle what goes into the model, where it is processed and what is kept.
Automated decisions from December 2026
Privacy law changes passed in 2024 require a privacy policy to disclose when a computer program uses personal information to make, or substantially help make, decisions that significantly affect individuals. We flag any workflow that does this while scoping, so the disclosure is written before it is due.
Guidance for AI Adoption (AI6)
The National AI Centre published its Guidance for AI Adoption on 21 October 2025: six essential practices, known as AI6, which replaced the ten guardrails of the Voluntary AI Safety Standard. It is voluntary, but it is what a board, an auditor or a large customer will ask about. More on AI governance in Australia.
How we work with Australian teams from Auckland
The method is the one we use in New Zealand: pick one recurring workflow per person, time it, build it into a Claude skill with the person who does it, teach the team in the room, and time it again.
- The same working daySydney and Melbourne are about two hours behind Auckland, so calls, reviews and fixes happen in your business hours, not overnight.
- In person by arrangementWe come to Sydney and Melbourne for the training day, workshops and the parts of an engagement that need the team in one room.
- Remote between visitsBuilding, reviewing and adjusting workflows runs remotely, on the tools and licences your team already uses.
- Training in your officePeople learn on their own work, on their own screens. That is why the training happens where the team sits, not on a webinar.
The same work as in New Zealand
Australian teams get the same three pieces of work, scoped for Australian data and obligations.
Claude training
Everyone set up and one real workflow each built, in a day in your office in Sydney, Melbourne or wherever the team sits.
AI transformation
Workflows built into the real work, a team trained to run them, and adoption measured against the baseline the team timed.
A second brain
The knowledge of the business written down where people and AI can both reach it, so it stops living in one person's head.
When to hire a Sydney-based firm instead
If the work needs a consultant in your office several days a week for months, or a large team on site at once, hire a firm in your city. If you need a few workflows running, the people who do the work trained to run them, and a partner who stays on between visits, that is what we do. For the full picture of how we work across both countries, see our AI consulting services.
Frequently asked questions
Do you have AI consultants in Sydney and Melbourne?
We are based in Auckland and work with Australian teams remotely, with in-person visits to Sydney and Melbourne by arrangement for training days, workshops and the on-site parts of an engagement. Sydney and Melbourne are about two hours behind Auckland, so the working day overlaps almost entirely.
Can you keep our data in Australia?
Yes. Where the data needs it, we deploy Claude so that every inference runs in an Australian region on infrastructure you control, rather than in a vendor's own cloud. Our Caregroup deployment works this way: Claude on every seat, every inference sent to Amazon Bedrock in Australia through the practice network's own gateway on AWS.
Which Australian rules does an AI project need to consider?
For most businesses, the Privacy Act 1988 and the Australian Privacy Principles, which apply to personal information whether a person or a model handles it. The National AI Centre's Guidance for AI Adoption sets out six essential practices; it is voluntary, but it is the reference a board or a customer will ask about. Sector rules, such as health records or financial services obligations, sit on top.
What changes for automated decisions in December 2026?
Privacy law changes passed in 2024 require a privacy policy to say when a computer program uses personal information to make, or substantially help make, decisions that could significantly affect an individual. If an AI workflow does that, the policy has to say so. We flag those workflows when we scope the work so the disclosure is written before it is due.
Is the work different for Australian teams?
The method is the same: one recurring workflow per person, timed, built into a Claude skill with the person who does it, taught in the room, and timed again. What changes is the scoping: Australian privacy obligations, where the data is processed, and which sector rules apply are settled before anything is built.
Do you work outside Sydney and Melbourne?
Yes. Brisbane and the rest of Australia work the same way: remote by default, in person by arrangement for the training and the parts of the engagement that need the team together.
When should we hire a Sydney-based firm instead?
If the work needs someone in your office several days a week for months, or a large team on site at once, a firm based in your city is the better fit. If you need a small number of workflows running, people trained to run them, and a partner who stays on remotely between visits, that is the work we do.
Find out what is worth building first
The AI Readiness Check is free and ends in a ranked list of where AI earns its keep in your business, including, where it applies, the answer that it does not.