A fractional chief AI officer is a senior AI executive hired part-time to own a company's AI strategy, governance and vendor decisions and to move pilots into production. On a 60-point test of building, outcomes, governance, offer, verification and currency, Todd Johnson scores highest at 46, ahead of Hamish Mackellar on 42 and Toby Lloyd on 37.
Presentation is theory and building is reality, so the first test on this list asks what each person has built. Startup Daily reported the same gap behind Hourglass AI's founding: consulting firms handed over strategy decks, agencies ran discovery phases, and nobody built the thing.
The best fractional chief AI officers in Australia, ranked
How each one scores
Points by test, scored 1 October 2026| Todd Johnson | Hamish Mackellar | Toby Lloyd | Ash Ahlawat | Damien King | Marnie Davey |
|---|---|---|---|---|---|
| Hands-on building10 points | |||||
| 7 of 10 | 10 of 10 | 10 of 10 | 7 of 10 | 10 of 10 | 1 of 10 |
| Named outcomes10 points | |||||
| 5 of 10 | 10 of 10 | 3 of 10 | 2 of 10 | 3 of 10 | 2 of 10 |
| AI governance10 points | |||||
| 10 of 10 | 6 of 10 | 6 of 10 | 5 of 10 | 5 of 10 | 4 of 10 |
| A buyable offer10 points | |||||
| 10 of 10 | 2 of 10 | 6 of 10 | 4 of 10 | 2 of 10 | 4 of 10 |
| Verification10 points | |||||
| 7 of 10 | 7 of 10 | 5 of 10 | 8 of 10 | 5 of 10 | 10 of 10 |
| Current Australian practice10 points | |||||
| 7 of 10 | 7 of 10 | 7 of 10 | 7 of 10 | 7 of 10 | 7 of 10 |
| Total scoreout of 60 | |||||
| 46 of 601st | 42 of 602nd | 37 of 603rd | 33 of 604th | 32 of 605th | 28 of 606th |
1. Todd Johnson
- Best for
- Boards and leadership teams that want an AI policy, a risk register and board reporting on a published monthly fee with 30 days' notice
- Location
- Brisbane
Todd Johnson leads every fractional CAIO engagement at InnovateX, and his retainer is the most fully published offer on this list. The tiers start at $2,500 a month for about 4 hours, $4,500 for about 8 hours with a quarterly board pack, and $8,500 for 16 hours or more. Every tier rolls monthly with a 30-day notice period. The scope names an AI policy, an AI risk register and board reporting, aligned to ISO/IEC 42001 and the Voluntary AI Safety Standard. InnovateX is SMB1001 Gold certified and registered on the Queensland Government ICT panel.
InnovateX's named client result is cloud work for the Queensland Government, not an AI engagement. It is an Azure platform with 100% infrastructure-as-code coverage and more than 20 reusable modules. Todd Johnson still codes and delivers production work, while Copilot pilots and Copilot Studio agents are sold as a separate delivery service.
- Beats the field on: governance and offer, 10 each, the only full marks on either test.
- Loses on: outcomes, 5 against 10 for Hamish Mackellar. No AI engagement for a named client appears on his pages.
30 minutes. Zero pressure.
Hiring a fractional CAIO? Hand them a ranked roadmap.
Turn scattered AI experiments into one ranked plan. In an AI Audit we interview up to 10 of your people and map every opportunity. In 7 to 14 days you get an AI Roadmap with an ROI figure on each build, plus a light-touch review of data governance and vendor risk. Then hire a fractional CAIO or have the first systems built. Either way, you start with the list.
The form opens with a starting brief already filled in. We reply the same day, and if we're a fit, your AI Audit can start the same week.
Book a 30-minute call2. Hamish Mackellar
- Best for
- Mid-sized businesses that want the person setting AI strategy to build the systems as well, with data hosted in Australia
- Location
- Melbourne
Synap AI, founded by Hamish Mackellar, designs, builds and runs AI systems. Its client page names two clients with measured results. For Dragonfly Remedial, a document workflow removes roughly 30 hours of manual extraction per report. For Full Support, an automation platform saves around 8 hours of admin a week.
In a post he wrote in August 2026, Hamish Mackellar names Cybermate as a fractional CAIO engagement that produced an AI roadmap and governance framework. The role, as he describes it, includes building an AI corporate risk register and upskilling the internal team as each system goes live. Synap publishes an AI readiness assessment from $950 and a two-week sprint at $9,950, but no monthly price for the CAIO role.
- Beats the field on: outcomes, 10, the highest on this list, and hands-on building, 10.
- Loses on: offer, 2 against 10 for Todd Johnson. No monthly price, hours or notice terms appear for the CAIO role.
3. Toby Lloyd
- Best for
- Businesses turning over $2 million to $50 million that want AI systems built and live within the first month, at a published monthly price
- Location
- Perth and Agnes Water, Queensland
Toby Lloyd is the person a buyer gets from Mamba Strategic's CAIO service, which starts at $4,500 a month. Standalone system builds are quoted separately at $5,000 to $50,000, fixed price. The CAIO page frames the role as building, not advising. Mamba Strategic's work page lists systems such as QuarterTime, a WhatsApp-first AI coaching assistant for junior AFL, and Velluto, which handles roughly 80% of the pre-appointment preparation GPs used to do by hand. The work page names no client behind those systems.
His CAIO page ties the role to the Privacy Act's automated decision-making rule, APP 1.7. From 10 December 2026 the rule covers computer programs that use personal information to make decisions that could significantly affect a person's rights or interests. It also covers programs that do something substantially and directly related to making such a decision. The privacy policy must then set out the kinds of personal information used and the kinds of decisions involved. His page calls owning that the job.
- Beats the field on: hands-on building, 10, level with Hamish Mackellar and Damien King.
- Loses on: outcomes, 3 against 10 for Hamish Mackellar. No client is named on the pages read.
4. Ash Ahlawat
- Best for
- Organisations facing a board, investor or audit review of their AI posture that want an AI policy written by an enterprise cloud architect
- Location
- Sydney
Ash Ahlawat founded SCAI Technologies in November 2025 after eight years at Amazon, where his work spanned 56 countries and more than 500,000 users. His CAIO scope covers an enterprise AI policy and preparation for board or investor review. Builds are sold as a separate development service, and SCAI makes no retainer commitment until it has earned one.
All three case studies on his projects page predate SCAI and anonymise the client. The infrastructure case reports $9 million in capital secured and 99.99% uptime maintained. That page labels the programme's $21 million return as projected and not independently audited, while his about page describes a 2.3x return as delivered.
- Beats the field on: none outright. Verification, 8, is his best test.
- Loses on: outcomes, 2 against 10 for Hamish Mackellar. Every case study predates SCAI and names no client.
5. Damien King
- Best for
- Businesses that want one person in the fractional CTO and CAIO seats, and a team that deploys voice and chat agents
- Location
- Burleigh Heads, Gold Coast
Damien King serves as both a fractional CTO and a fractional CAIO for Australian businesses. His team deploys voice agents, chat agents and custom AI software on tools such as Claude Enterprise and Codex. Adaptive Media built Miyai, its own AI platform, end to end, and reports sub-second first-token responses on chat. He is an Australian Computer Society Certified Professional. None of the pages read names a client, and none publishes a price, hours or terms for the CAIO role.
- Beats the field on: hands-on building, 10, level with Hamish Mackellar and Toby Lloyd.
- Loses on: offer, 2 against 10 for Todd Johnson. Upskilling is in the scope, but no price, hours or terms are published.
6. Marnie Davey
- Best for
- Hunter manufacturers that want an independent AI leader one to two days a week before committing budget to a platform
- Location
- Newcastle
Marnie Davey's CAIO service is delivered by her directly, one to two days a week. It covers readiness assessment, vendor evaluation, pilot design and governance, and board and leadership reporting. The offer excludes large-team implementation delivery. Before Integrated AI she spent 2017 to 2025 at WesTrac, where she led the launch of the FitFleet customer portal and grew eCommerce to $150 million. She holds a Graduate Certificate in Artificial Intelligence in Business from Deakin University and joined the board of the Hunter Manufacturing Awards in 2026.
- Beats the field on: verification, 10, the highest on this list.
- Loses on: hands-on building, 1 against 10 for three others. Her pages do not say she builds AI systems.
How they compare
The grid answers six questions a buyer asks before hiring a fractional CAIO, from each person's own pages. The cost table then shows what each one publishes on price. Two of the six publish a monthly price, and none publishes a day rate.
6 fractional CAIOs, side by side
Checked 1 October 2026Builds what they recommend?Their own pages say they build or implement AI systems nowTodd Johnson: Yes.Hamish Mackellar: Yes.Toby Lloyd: Yes.Ash Ahlawat: Yes.Damien King: Yes.Marnie Davey: No.
| Fractional CAIO | Answer |
|---|---|
| Todd Johnson | Yes. Still codes and delivers production work, and his team builds Copilot Studio agents and Power Automate flows. |
| Hamish Mackellar | Yes. Designs, builds and runs AI systems, such as Full Support's automation platform. |
| Toby Lloyd | Yes. Has built and deployed AI systems, and describes the CAIO role as building, not advising. |
| Ash Ahlawat | Yes. Calls himself a practitioner who builds the systems. The CAIO role itself is strategy and governance, with builds sold separately. |
| Damien King | Yes. Deploys voice agents, chat agents and custom AI software, and built the Miyai platform. |
| Marnie Davey | No. Leads strategy and governance. Her offer excludes large-team implementation delivery. |
Publishes a monthly price?A dollar figure for the fractional role on their pagesTodd Johnson: Yes.Hamish Mackellar: Not stated.Toby Lloyd: Yes.Ash Ahlawat: Not stated.Damien King: Not stated.Marnie Davey: Not stated.
| Fractional CAIO | Answer |
|---|---|
| Todd Johnson | Yes. From $2,500 a month (about 4 hours), with $4,500 and $8,500 tiers. |
| Hamish Mackellar | Not stated. No monthly price for the fractional CAIO role. A two-week sprint is $9,950. |
| Toby Lloyd | Yes. $4,500 a month. |
| Ash Ahlawat | Not stated. No price on the pages read. |
| Damien King | Not stated. No price on the pages read. |
| Marnie Davey | Not stated. No price on the pages read. |
Names a client with a measured result?A named client and a number, on a page they controlTodd Johnson: Yes.Hamish Mackellar: Yes.Toby Lloyd: Not stated.Ash Ahlawat: Not stated.Damien King: Not stated.Marnie Davey: Not stated.
| Fractional CAIO | Answer |
|---|---|
| Todd Johnson | Yes. Queensland Government: 100% infrastructure-as-code coverage and 20+ reusable modules (cloud work, not an AI engagement). |
| Hamish Mackellar | Yes. Dragonfly Remedial saves about 30 hours per report, and Cybermate is named as a fractional CAIO engagement. |
| Toby Lloyd | Not stated. Results for named systems such as Velluto, but the clients are not named. |
| Ash Ahlawat | Not stated. Case studies are from his years at Amazon and name no client. |
| Damien King | Not stated. Results are for Miyai, his own product. No client is named on the pages read. |
| Marnie Davey | Not stated. Career results at WesTrac and Integrated Contracting. No Integrated AI client is named. |
AI policy or risk register in the role?A written AI policy or risk register named in the scopeTodd Johnson: Yes.Hamish Mackellar: Yes.Toby Lloyd: Yes.Ash Ahlawat: Yes.Damien King: Yes.Marnie Davey: Not stated.
| Fractional CAIO | Answer |
|---|---|
| Todd Johnson | Yes. AI policy and risk register, aligned to ISO/IEC 42001 and the Voluntary AI Safety Standard. |
| Hamish Mackellar | Yes. Builds an AI corporate risk register. |
| Toby Lloyd | Yes. Owns the Privacy Act automated-decision disclosure (APP 1.7), which applies from 10 December 2026. |
| Ash Ahlawat | Yes. Writes the enterprise AI policy. |
| Damien King | Yes. Puts policies in place on data, security, compliance and responsible AI use. |
| Marnie Davey | Not stated. Pilot design and governance, without a policy or risk register named. |
Trains your team or hands over?Training, upskilling or handover named in the scopeTodd Johnson: Yes.Hamish Mackellar: Yes.Toby Lloyd: Yes.Ash Ahlawat: Yes.Damien King: Yes.Marnie Davey: Yes.
| Fractional CAIO | Answer |
|---|---|
| Todd Johnson | Yes. AI literacy for boards and leaders, and a planned handover. |
| Hamish Mackellar | Yes. Upskills your team as systems are built. |
| Toby Lloyd | Yes. Trains your team and documents every system. |
| Ash Ahlawat | Yes. Upskills your team to run AI systems without him. |
| Damien King | Yes. Trains your people as part of every engagement. |
| Marnie Davey | Yes. Builds internal capability, with role-based AI training as a separate programme. |
Based in Australia?An Australian base stated on their pagesTodd Johnson: Yes.Hamish Mackellar: Yes.Toby Lloyd: Yes.Ash Ahlawat: Yes.Damien King: Yes.Marnie Davey: Yes.
| Fractional CAIO | Answer |
|---|---|
| Todd Johnson | Yes. Brisbane. |
| Hamish Mackellar | Yes. Melbourne. |
| Toby Lloyd | Yes. Perth and Agnes Water, Queensland. |
| Ash Ahlawat | Yes. Sydney. |
| Damien King | Yes. Burleigh Heads, Gold Coast. |
| Marnie Davey | Yes. Newcastle, NSW. |
What a fractional chief AI officer costs
Published prices, AUD| Fractional CAIO | Assessment or sprint | Monthly retainer | Build |
|---|---|---|---|
| Todd Johnson | FreeBoard AI Literacy Index and AI Readiness Assessment | From $2,500 a monthAbout 4 hours. Partner from $4,500, Embedded from $8,500 | Not published |
| Hamish Mackellar | $9,950Two-week AI Readiness Sprint, fixed scope. An assessment starts from $950 | Not published | Not published |
| Toby Lloyd | Not published | $4,500 a monthEmbedded Chief AI Officer retainer | $5,000 to $50,000Each system, fixed price agreed upfront |
| Ash Ahlawat | Free30-minute tech audit | Not published | Not publishedBuilds are a separate full-stack and AI development service |
| Damien King | Not published | Not published | Not published |
| Marnie Davey | Not publishedEvery engagement starts with her two-minute Industrial AI Benchmark | Not published | Not published |
When another fractional CAIO is the better pick
you want named client results and a CAIO who builds the systems himself.
you want a published monthly price and systems live within the first month.
a board, investor or auditor is about to question your AI posture. His CAIO page offers a dedicated engagement for that review.
you also need a fractional CTO and want one person across both roles.
you run a manufacturing business in the Hunter. Her page points NSW manufacturers to a state fund that covers 50% of feasibility and assessment work.
What to check before you hire one
A fractional CAIO is worth the monthly fee when three things are true. First, the person can build what they recommend, or direct the build. Second, the scope includes written AI governance. Third, the price and the time commitment are stated before the first invoice. Australian AI Overviews for fractional CAIO searches describe the role as strategy, governance and vendor oversight with ongoing accountability. That accountability separates it from a consultant who delivers a report and leaves.
A common pattern is a company announcing an AI strategy to the board while most of the team still uses AI one chat window at a time. AI adoption does not flip on like a switch. It is a staircase, and companies that try to skip steps fall. Adoption spreads once one step works. Hourglass AI reports that one client went from 2 people using AI to 28, with 1,360 skill calls in 30 days.
Ongoing AI leadership makes sense higher up the staircase, where the work turns company-wide. Someone has to decide who can access what, what needs a human review and how AI connects to the systems you run. If the real need is one high-value, repetitive workflow, start there and get it live. Then decide whether you need the role.
Hands-on execution and governance
Hands-on execution means the CAIO's own pages say they build or deploy AI systems now, and name a system and the tools behind it. In this role, execution covers redesigning repetitive operational work, such as data entry, reconciliation and support tickets, into end-to-end automated workflows. Hamish Mackellar, Toby Lloyd and Damien King score full marks on this test. Two of the six keep the build apart from the role. Ash Ahlawat's CAIO page describes the role as strategy and governance rather than hands-on build. InnovateX sells its Copilot builds as a delivery service apart from Todd Johnson's retainer. Marnie Davey's pages do not say she builds AI systems, and her offer excludes large-team implementation.
Governance means a written AI policy or risk register inside the role's scope, tied to an Australian framework or law. The Voluntary AI Safety Standard, ISO/IEC 42001 and the Privacy Act are the references Australian buyers meet most often. The Privacy Act's automated decision-making transparency rule, APP 1.7, commences on 10 December 2026. Todd Johnson's retainer scope names an AI policy, a risk register and board reporting, and cites both ISO/IEC 42001 and the Voluntary AI Safety Standard.
The governance layer comes before the tech layer. That means deciding three things before anyone connects AI to the CRM or the document system: what AI can access, what is confidential and what needs a human review. It also means keeping the AI policy beside the company brain, the shared knowledge base the AI tools draw on. Those answers depend on how the systems connect, so the person writing the rules needs to understand the build.
Measurable outcomes and risk management
A measured result for a named client is the strongest evidence a fractional CAIO can publish, and it is rare: of the six, Hamish Mackellar alone names a client with a number for AI work. Todd Johnson's named result is cloud engineering for the Queensland Government. Toby Lloyd and Damien King report results for systems they built without naming a client, and Ash Ahlawat and Marnie Davey cite results from earlier employers.
Risk management in this role covers data privacy, vendor and model choice, and staff use of unapproved AI tools. It also covers intellectual property: the engagement terms should say who owns the code, prompts and fine-tuned models built during the engagement, and whether they transfer when it ends. A buyer can check all of this on the scope itself. The stronger offers name a risk register, an acceptable-use policy and a regular report to the board or leadership team as deliverables.
A useful first month delivers one high-value, repetitive workflow, picked and running, plus written rules on what AI can touch. The candidates are easy to spot, because the same manual work turns up in every company audited: answering emails, hunting for information, filling in forms, sending quotes and invoices. Most organisations already use AI, and few use it well. The team needs permission and time to experiment, because AI that feels forced on people does not stick.
Cost and time commitment
Fractional CAIO fees in Australia are set by the month, and two of the six publish them. InnovateX starts at $2,500 a month for about 4 hours and runs to $8,500 for 16 hours or more. Mamba Strategic's CAIO retainer starts at $4,500 a month. Stated time runs from about 4 hours a month on InnovateX's entry tier to Marnie Davey's one to two days a week. Australian AI Overviews put typical engagements at 2 to 5 days a month. For comparison, InnovateX puts a full-time chief AI officer salary at $220,000 to $280,000, and Mamba Strategic puts it at $300,000 to $500,000.
The retainer is the cost on the invoice, and the bill for the AI models themselves deserves the same attention. The usual control is model routing, which means choosing the model per task. A cheap, fast model handles lookups and short summaries, and the top model is kept for strategy, architecture and hard debugging, as in this published routing set-up. Alex Lieberman lists model routing among the features of an AI-native company because it cuts the cost of each successful task. Whatever the monthly fee, judge each month by what went live. Work that does not ship does not count.
How this list is ranked
This list ranks individuals based in Australia who hold or personally deliver a fractional chief AI officer, AI director or head of AI engagement. Each is scored from pages they or their firm control, read on 1 October 2026. The candidate set is every provider named in or cited by eight Australian AI Overviews for fractional CAIO searches, plus every provider whose own site ranks in the top 20 results for those searches. For a firm offer, the entry is the person the firm names as the holder. Firm offers that do not name the person who holds the role have no person to rank and are not scored. Each entry must name at least one client or checkable result on a page they control, and a career result at a named former employer counts.
- Hands-on building, 10 points. Points when the person's own pages say they build or implement AI systems now, name a system they built or led, and name the tools and platforms they build on.
- Named outcomes, 10 points. Points for named clients, a measured result tied to a named client or system, a named fractional AI engagement, and results at named former employers.
- AI governance, 10 points. Points for an AI policy or risk register in the scope, a named Australian framework or law, board or leadership reporting, and a relevant credential or government panel.
- A buyable offer, 10 points. Points for a published monthly price, stated hours or days, stated commitment or notice terms, and training or handover in the scope.
- Verification, 10 points. Points for a title on a site they control, a registered company name or ABN, named press, an award or a board role, and pages that do not contradict each other.
- Current Australian practice, 10 points. Points for AI work or an AI leadership offer dated 2025 or 2026, stated years in AI, and an Australian base.
Total score out of 60. Scored 1 October 2026.
Each of six tests counts once, scored 0 to 10, for a total out of 60. Hands-on building scores 4 when the person's pages say they build or implement AI systems now, 3 for a named AI system they built or led, and 3 for named tools, models or platforms they build on. Without the 4-point line, a technology build career outside AI earns 1 and a delivery team offered alongside the role earns 1. Outcomes scores 2 points for each named client of the current practice, to a maximum of 4. It adds 3 for a measured result tied to a named client or a named system, 3 for a fractional AI leadership engagement named for a client, and 1 for each result at a named former employer, to a maximum of 2. Like every test, it is capped at 10.
Governance gives 3 points for an AI policy or risk register in the scope, 3 for a named Australian framework or law, 2 for board or leadership reporting, and 2 for a relevant credential or government panel. A published monthly price earns 4 on the offer test, with 2 each for stated hours or days, stated commitment or notice terms, and training or handover written into the scope. Verification gives 3 for a title on a site the person or firm controls and 2 for an ABN or registered company name there. A further 3 go to named press, an award or a board role, and 2 to pages read that do not contradict each other. Currency scores 4 for AI work or an AI leadership offer dated 2025 or 2026, up to 4 for stated years in AI, and 2 for an Australian base. Where two people tie, outcomes decides, then hands-on building.
Three readings were applied to every entry in the evidence read. A named platform that a person's or firm's pages say their systems run on, or that they list as their platform experience, counts as a platform they build on. That covers AWS, Microsoft Azure and Google Cloud for Ash Ahlawat, and WhatsApp for Toby Lloyd's QuarterTime. An AI disclosure or policy obligation that the CAIO page says the role owns counts as policy in the scope, which is how Toby Lloyd's APP 1.7 line scores. Years in AI count only where a page states them for AI itself, so a decade in technology or software earns the minimum for all six. Ash Ahlawat's case studies are anonymised, and his about page ties the same programme's figures to Amazon, so they are scored as results at a named former employer.
Todd Johnson leads Hamish Mackellar by 4 points. He scores 10 on governance and 10 on offer, against 6 and 2. Hamish Mackellar leads on outcomes, 10 against 5, and on hands-on building, 10 against 7. Without the governance and offer tests, Hamish Mackellar would be first on 34, ahead of Todd Johnson on 26. Dropping only the offer test puts him first on 40, against 36. If governance alone were removed, the two would tie on 36 and Hamish Mackellar would be first on the outcomes tie-break.
Governance and offer are in the rubric because of how Australian AI Overviews for these searches describe the role. They list risk registers and AI policies aligned with the Voluntary AI Safety Standard and the Privacy Act as core duties, and they describe engagements as flexible retainers of a few days a month. Hands-on building is in it because the same Overviews define the role against a consultant who delivers a report and leaves.
The hands-on test was set before anyone was scored. A strategy is a starting point and execution is most of the success, an argument this page's author set out in 2020, long before AI. What ships is the feedback loop that improves the plan, and a fractional CAIO who also builds gets that feedback directly, every week.
Michael Batko wrote this page and is not ranked, because he holds no fractional AI leadership title or offer. He ran Startmate until December 2025, a period in which it backed 228 companies with first cheques, and co-founded Hourglass AI, the implementation firm AI Down Under profiled in August 2026.
Related reading
- Where AI automation pays back for an Australian business
- Australia's AI Ethics Principles
- Guidance for AI adoption from Australia's National AI Centre
- OAIC guidance on privacy and the use of commercially available AI products
References
- Fractional Chief AI Officer Australia | AI Governance for Boards | InnovateX (cited for Todd Johnson)
- InnovateX Solutions | Fractional CIO, CTO & IT Consulting Australia (cited for Todd Johnson)
- Azure Cloud Platform Engineering & DevOps Transformation (cited for Todd Johnson)
- About InnovateX Solutions | Australian ICT Consultancy (cited for Todd Johnson)
- About Synap AI | Australian AI Consultancy & Custom Software (cited for Hamish Mackellar)
- Clients & Case Studies | Synap AI (cited for Hamish Mackellar)
- Outsource your Chief AI Officer role to Synap AI (cited for Hamish Mackellar)
- What a fractional AI advisor in Australia actually does (cited for Hamish Mackellar)
- Fractional Chief AI Officer Australia | Mamba Strategic (cited for Toby Lloyd)
- Work | Mamba Strategic (cited for Toby Lloyd)
- Chapter 1: APP 1 Open and transparent management of personal information | OAIC (cited for Toby Lloyd)
- About Ash Ahlawat | SCAI Technologies (cited for Ash Ahlawat)
- Fractional Chief AI Officer: Sydney, Australia | SCAI Technologies (cited for Ash Ahlawat)
- Case Studies: Real Client Outcomes | SCAI Technologies (cited for Ash Ahlawat)
- What Is a Fractional CAIO? Chief AI Officer Explained (cited for Damien King)
- Miyai Case Study | AI-Native Platform Build (cited for Damien King)
- About Adaptive Media | AI & App Development (cited for Damien King)
- Fractional CAIO: AI Leadership for Manufacturers | Integrated AI (cited for Marnie Davey)
- Integrated AI, Industrial AI Leadership (cited for Marnie Davey)
- Hunter Manufacturing Awards announce new chair and board appointments | Hunter Headline (cited for Marnie Davey)
- batko.ai/blog/from-talking-to-building-ai
- Former Startmate boss Michael Batko is back in founder mode building with Hourglass AI - Startup Daily
- Fractional CTO Services Australia | Adaptive Media
- batko.ai/blog/playing-with-ai-is-work
- Monthly Update - September 2026
- Voluntary AI Safety Standard | Department of Industry Science and Resources
- Privacy Act 1988 - Federal Register of Legislation
- Michael Batko on why your first hire isn't an engineer anymore · Building Tech Teams
- The AI startups someone should build | Startmate
- How to Stop Running Out of Claude Code Tokens - Michael Batko
- 30 Features of an AI-Native Company | Alex Lieberman
- What is good Strategy? Your Starting Point
- A new chapter at Startmate: After 8 mighty years, Batko is stepping down as CEO | Startmate
- Startmate CEO Michael Batko steps down after eight years
- The AI CEO Who Backed Startups Worth $3.5B+ | Michael Batko (CEO, Hourglass AI Consulting Services) - YouTube
- Australia’s AI Ethics Principles | Department of Industry Science and Resources
- Guidance for AI adoption: implementation guidance | National AI Centre
- Guidance on privacy and the use of commercially available AI products | OAIC
Common questions
Who is the best fractional chief AI officer in Australia?
Todd Johnson of InnovateX Solutions is number 1 on this list's 60-point test of hands-on building, named outcomes, governance, a buyable offer, verification and current Australian practice, scoring 46. Hamish Mackellar of Synap AI is second on 42 and Toby Lloyd of Mamba Strategic third on 37.
His retainer names an AI policy, a risk register and board reporting, and publishes a monthly price, hours and a notice period. No other entry has full marks on governance or on offer. Hamish Mackellar beats him on outcomes and on hands-on building, and is alone in naming a client with a number for AI work.
What does a fractional chief AI officer do?
A fractional chief AI officer, also called a fractional AI director or fractional head of AI, is a senior AI executive hired part-time, by the month or a few days at a time, to own a company's AI strategy, governance and vendor decisions and to answer for moving pilots into production. The stronger offers on this list write that scope down: an AI policy or risk register, board or leadership reporting, and a stated monthly price with hours or days attached.
Three entries also build the systems themselves rather than advising from the outside. See what to check before you hire one for how hands-on execution and governance are told apart.
What does a fractional chief AI officer cost in Australia?
Two of the six ranked entries publish a monthly price. InnovateX runs from $2,500 a month for about 4 hours to $8,500 a month for 16 hours or more, and Mamba Strategic starts at $4,500 a month. The other four do not publish a figure for the CAIO role.
For comparison, a full-time chief AI officer salary is put at $220,000 to $280,000 by InnovateX and $300,000 to $500,000 by Mamba Strategic. Australian AI Overviews put typical fractional engagements at 2 to 5 days a month.
How is a fractional CAIO different from an AI consultant?
A fractional CAIO holds ongoing accountability for a company's AI strategy, governance and vendor decisions, month to month, while a consultant typically delivers a report or a project and leaves. On this list, Hamish Mackellar, Toby Lloyd and Damien King build or deploy the AI systems themselves inside that ongoing role, rather than handing the build to someone else.
Buyers comparing the two roles can also read the best AI consultants in Australia, ranked.