Home
You’re currently viewing our content for All Regions.
You can switch to a different region:
June 24, 2026
As with all transformative technology, market attention naturally fixates on the early pace-setters – the sharp end of the industry where businesses are building on the promise of agentic AI at speed and scale.
At the bleeding-edge, the AI air is thin. Deployments are fast, use cases are precise and models are mature.
Whether kicking on or catching up from the laggard-last, Australia stands at an inflection point in AI adoption. After years of experimentation, pilots and cautious exploration, organisations are shifting from curiosity to capability-building.
But behind the customer, the use case, the model and the platform is an ecosystem.
And Google Cloud has just fired the agentic AI starting gun on its high-value network of partners.
“Google Cloud’s partners are already leaders in agentic AI development and deployment, and have become important channels for distributing AI technologies,” said Kevin Ichhpurani, President of Global Partner Ecosystem at Google Cloud. “We are committed to offering customers the most AI-capable partner ecosystem in the industry.”
Rather than pursuing a strategy built purely on partner volume, Google Cloud has long focused on cultivating an ecosystem capable of delivering meaningful customer outcomes. The result is a network of partners that combines deep technical expertise, industry knowledge and a shared commitment to innovation.
Unveiled at Cloud Next ’26 in April, the technology giant has rolled out a $750 million fund to deliver new resources and incentives to partners to help accelerate joint customer transformations with agentic AI.
The fund – available for global consulting firms, systems integrators, software partners and channel partners – is designed to support AI value identification, agentic AI prototyping, agent building and deployment, up-skilling, and teams of embedded Google Forward Deployed Engineers (FDEs).
“With this expanded funding, we will be able to dedicate new resources and technology to support our partners as they accelerate our mutual customers’ agentic AI journeys,” Ichhpurani added.
In other words, this is the arming and mobilising of a 120,000-member partner ecosystem.
To put those numbers into context, Google Cloud’s ecosystem of system integrators already offer more than 330,000 experts trained on implementing Google AI for customers, and 95% of the top 20 and over over 80% of the top 100 SaaS companies use Gemini models.
One such partner primed and ready to capitalise on the agentic AI opportunity is Aviato Consulting, recently honoured as 2026 Google Cloud Partner of the Year across Australia and New Zealand.
“We’re leveraging Google Cloud’s new $750 million partner fund to co-develop minimum viable products (MVPs) on real production data, utilising embedded FDEs to prove the business value before full-scale deployment,” said Ben King, Founder and CEO of Aviato.
“This move to more of an Outcome-as-a-Service (OaaS) model is very new from Google Cloud and more closely aligns with our approach. By funding our engineers to work with customers to deploy Google Cloud technology on specific business problems in customer environments, we can jointly accelerate time to value for businesses in Australia.
“The funding from Google Cloud will de-risk this approach meaning we can do more of this, without maxing out our pre-sales engineering budget.”
Aviato was recognised for “outstanding achievements” within the Google Cloud ecosystem, specifically for helping joint customers modernise infrastructure and accelerate digital transformation through specialised cloud consultancy. This is in addition to delivering “unique impact” to drive innovation and deliver measurable and scalable success for global enterprises.
“Our biggest value add is helping organisations transition from simply digitising existing processes to true ‘agent-ification’,” King added.
“We focus heavily on orchestrating models and tools using the newly announced Gemini Enterprise Agent Platform and the open Agent2Agent (A2A) protocol, which allows diverse agents from multiple vendors to seamlessly collaborate.”
Security is also paramount with Aviato integrating governance via tools such as AI Defence Plane by Check Point and AI Application Protection Platform (AI-APP) by Wiz to monitor runtime behaviour.
As noted by King, channel economics are undergoing a “paradigm shift” from traditional technical rate cards to outcome-focused solutions. In the local market, Aviato is actively embracing OaaS models with the aim of delivering “guaranteed business results” rather than just billing for integration hours.
“With an estimated $20 trillion total addressable market across 120 job families, the ecosystem will increasingly rely on cross-platform interoperability,” King outlined.
“We’re already seeing native integrations between Google Cloud’s infrastructure and enterprise platforms such as Salesforce and SAP, facilitated by the A2A protocol and easily discoverable through the Gemini Agent Gallery. This frictionless ecosystem will be the main driver as customers rapidly scale into full production.”
Unlike the previous cloud cycle, the AI opportunity isn’t a “won and done” proposition in Australia. That’s the view of Tom Bernadou, Global Google Alliance Lead at NCS Group.
“Cloud followed a predictable path – assess the infrastructure, pitch the business case, migrate and hand over the keys,” Bernadou detailed.
“Winning with any AI-enabled work means ensuring ongoing adoption. Agents have to learn and evolve and so our service models are shifting to match that. It’s not just deploying technology, our incentives align with long-term success and the performance of that AI inside a client’s business.”
From a partner value standpoint, NCS – who acquired Riley, the Google Cloud partner business that Bernadou founded in 2021 – is delivering technology expertise paired with domain knowledge, now viewed as the winning customer formula in agentic AI.
“You have to go end-to-end, however – and the home runs come from industry expertise,” Bernadou added.
“In retail and telco, we have seen our domain understanding become mandatory. To effectively orchestrate a workflow, we need to understand the nuance of the business process.
“Having deep industry knowledge helps us to frame how the models and tools should be used, and from there everything else follows. For example, in telco, it’s not just building an FAQ agent but instead delivering one that can manage mobile plans or technically troubleshoot outages for customers.”
This approach is backed up by industry data. According to Moxie Research – AI Outlook: Australia 2026 – the most important characteristics that Australian organisations seek when working with an AI partner are:
At Mantel, the specialist consultancy firm is focused on four key areas during the next 12 months.
“First, how agents learn – from SMEs, from customers, from their own failures – and how that feedback is captured both as process knowledge and as hyperpersonalisation,” shared Vihan Patel, Head of AI Solutions at Mantel.
“The end state we’re working towards is agents that self-heal and evolve in line with business needs, the way mature MLOps loops do for classical ML models today.”
Second, the unit economics of building agents. Mantel has scaled many leading customers with the question now anchored on how to bring the cost of each new complex agent down without sacrificing engineering quality – achieved through repeatable patterns and extracting the most out of agentic coding tools.
Third is focused on observability and governance in production, both for responsible AI reasons and for ROI measurement. The business is “actively exploring” how to automatically measure the ROI of production agents, so that signal can feed back into the business and inform what gets built next.
“Finally, adoption is a key challenge that all of our enterprise customers face,” Patel added.
“We’ve developed an initial offering to support both broad adoption and strategic investments. Refining this offering will be a focus, as we bring together agentic adoption patterns for products, engineering teams and business users.”
This approach is anchored on a commitment to add increased levels of value across the entire customer chain.
“Upfront, we’re doing process mapping, redesign and user experience design – we’ve created some genuinely cutting-edge frameworks for designing effective agentic UI,” Patel continued.
“Our engineering experience across some of Australia’s largest and most heavily regulated enterprises means that we can execute at scale in a repeatable way. And throughout, we take client teams on the journey – capability uplift happens alongside delivery, because we don’t want to leave clients with something they can’t manage or with a dependency on us to keep it running.”
The business has also introduced two-speed thinking which aligns mass adoption with strategic investment in key use cases. The belief is that enterprises shouldn’t delay rolling out business facing agentic tooling while they build specific agentic use cases centrally.
“Having a change management program to ensure the value return on any agentic tooling is accelerated in parallel to building solutions with the engineering teams is the emerging best practice,” Patel said.
While Onigroup operate as a “deeply technical” partner of Google Cloud, this provider’s highest value-add is delivering strategic direction and vision.
“In a world of non-deterministic outputs, teams need to be comfortable with a certain level of fluidity,” said Darragh Murphy, Managing Director of Onigroup.
“We educate executives on how to manage this new paradigm. We act as the orchestrators – designing the workflows and ensuring the security and compliance frameworks are ‘Google-grade.’
“But more importantly, we provide the alignment layer. We ensure that the AI strategy isn’t a siloed IT project but a core business driver. We help teams move from the ‘fear of the unknown’ to the ‘mastery of the new,’ ensuring that adoption is both safe and culturally supported.”
According to Murphy, the rate of innovation coming out of Google Cloud is “staggering”. For Onigroup, this means moving away from traditional “one-and-done” implementation models towards continuous enablement and optimisation.
“As customers move from pilots to full-scale production, we see the emergence of ‘AI managed services’,” Murphy added.
“This isn’t just keeping the lights on; it’s the constant tuning of agents, model switching as better versions of Gemini are released, and evolving agentic workflows as business needs change.
“We’re positioning ourselves to meet customers where they are – whether they need a foundation for innovation or a partner to manage a complex, multi-agent ecosystem. Our future is in being the long-term innovation partner for businesses that refuse to be left behind.”
During the next 6-12 months, Australian businesses will transition through three key phases when prioritising key agentic AI initiatives.
That’s according to Moxie Research. This AI research is based on a total of 520 survey responses from IT decision-makers across Australia, generating more than 250 data points – over 30 hours of C-suite interviews with 28 influential leaders spanning 16 market sectors.
Phase 1
Phase 2
Phase 3
“We’re moving past the ‘play phase’ of GenAI,” observed Bernadou of NCS. “The market is shifting from simple chatbots towards agent workflows where AI doesn’t just talk, it takes action.”
Partnering with Google Cloud offers NCS access to superior models, more robust infrastructure and advanced tool use – and with that AI can now confidently reach into a core business system to execute a real task.
“I wouldn’t say we have ‘ambient’ (fully autonomous) deployments just yet,” Bernadou continued. “But as leaders become more confident in agent performance, it won’t be long before we see autonomous agents integrated into the real-time operations of Australian businesses.”
This shift in intent and investment is playing out at scale with agentic AI development moving into the mainstream across a range of industries, each housing unique use cases by business function.
“Businesses have gone from ‘wow’ to ‘how’ very fast,” noted King of Aviato.
“Agentic AI development has absolutely gone mainstream in Australia with Gemini Code Assist already in use by a lot of our customers. We’re seeing more autonomous systems, moving into mission critical workflows.”
In assessing current market readiness, King acknowledged that most businesses “plateau” around 20-50 agents.
“Why? Because businesses aren’t giving much thought about how they manage agents,” King explained. “They have extensive pipelines to deploy their code but possess no thought about how to manage agents – much the same as having an HR department, you need an HR department for agents.
“For example, AI requires a performance review every time a system prompt is updated or moved from Gemini 2.5 to 3.0. The business wants to know if that upgrade is going to improve the output, and justify the potential cost increase.”
As the National AI Plan sets out a pathway for Australia to be a ‘developer and adopter of trusted, world-class AI solutions’, creating the right environment for investment, innovation and adoption to flourish is mission-critical.
The organisations that succeed will be those that move beyond surface-level tools and build the underlying engines – data, people, processes and platforms – that turn AI from a conversation into capability, and from capability into measurable value.
“We’re well past the conceptual stage,” reiterated Patel of Mantel.
“Leading organisations are deploying agents across the business at both ends of the spectrum – low-code and no-code variants enabled by Gemini Enterprise, and complex, high-code solutions built on ADK. The conversation has moved from whether to deploy agents, to where to start and how to scale.
“For example, we have a team deployed to one large customer focusing on end-to-end adoption and the value of Google Cloud’s agentic technology. This includes an engineering team to deploy custom built, fully functional agents, deployed engineers to train and uplift business teams and change managers to help ensure adoption of low-code/no-code options.”
Meanwhile, Murphy of Onigroup acknowledged a “broad spectrum” of agentic AI readiness across Australia.
On one end, digital natives and early adopters are already moving beyond simple chatbots into realising the true capabilities of agentic AI – deploying autonomous agents on Google Cloud that can reason, use tools and execute multi-step tasks.
“These companies aren’t just ‘testing’ AI, they’re embedding it into their core value chain both for product development and within their products,” Murphy outlined.
On the other end, many traditional enterprises in the region still remain in the discovery phase. For those businesses, agentic AI remains conceptual as they grapple with the fundamental shift from deterministic software (where A always leads to B) to probabilistic / non-deterministic AI.
“Our role at Onigroup is to bridge this gap, helping realise that with Gemini Enterprise Agent Platform, the infrastructure to move from concept to production is already mature; the hurdle is now strategy, not just technology,” Murphy added.
Across Australia – and based on Moxie Research – 70% of organisations are already running internal AI working groups to shape strategy and drive adoption, with 12% under consideration.
The key AI benefits of cross-collaboration between business units are ranked as:
This offers a clear indication as to where businesses are prioritising agentic AI today, notably in the areas of customer service, internal productivity, decision support and industry-specific automation.
“We see priorities split between industries,” observed King of Aviato.
“Customer service and experience is the primary focus, with retail reporting huge uplift from customers using ‘concierge-like’ AI agents. For example, Macy’s built its ‘Ask Macy’s’ AI agent within four weeks and reported a 4.45x spend increase by shoppers using the platform.”
Demand for data agents is also increasing while conversational analytics is now viewed as a “big area of growth” – reliance on dashboards for key metrics will remain but the ability for executives to dive into the data at speed will allow businesses to gain competitive advantage in new way. This is taking shape through Google Looker, an enterprise platform for business intelligence and embedded analytics.
Building on this, Murphy of Onigroup highlighted that in Australia, current business focus in centred on “human-in-the-loop” decision support – using Google Cloud’s Gemini models to ingest massive datasets and provide agents with the reasoning capabilities to assist senior stakeholders.
“The driver here is simple,” Murphy continued. “Speed to insight is the new competitive moat.”
According to Moxie Research, the three domains proving consistently impactful as AI use cases in Australia today are:
From a decision-making standpoint, predictive models, real-time analytics and natural-language interfaces are giving leaders clearer visibility into risk, demand, staffing, asset utilisation and customer behaviour.
For many businesses, the shift is from having static dashboards to continuously learning systems that surface recommendations, test scenarios and anticipate issues before they become costly problems.
Based on Moxie Research, 55% of Australian businesses are already using AI to enhance strategy data and input with 24% actively spinning up agents to keep pace. This is a similar story for executive intelligence – 48% of organisations are already using AI agents for this purpose while 28% are in fast pursuit and “planning to use” within the next six months.
“Another big opportunity is the combination of domain knowledge and AI, which is why vertical-specific solutions seem to be the highest priority,” continued Bernadou of NCS.
“We have seen this first play out in retail and consumer sectors. We deployed a Google Cloud model tailored to industry specific outcomes which created a much stronger business case than delivering generic AI.
“When the goals were clear, the path to value was also. We’d hit a goal (an increase in revenue for example), and adoption and usage would naturally follow.”
At Mantel, use case development is “genuinely diverse” – such as internal HR agents serving hundreds of thousands of employees, financial reconciliations, invoice-to-pay, safety and frontline operations, through to customer-facing agentic commerce.
“Customer service is often where people expect to start but it’s frequently tangled up in legacy systems and commercial agreements that are slow to disrupt,” added Patel of Mantel.
“Even when there’s a top-down mandate, investment tends to flow towards the business units most engaged and open to evolving how they work. Those teams become the flagships, and the patterns and platforms they build end up forming the backbone of the broader program.”
In alignment, Murphy of Onigroup outlined that current business priorities are “heavily weighted” towards internal productivity and customer service and support.
“Early adopters initially deployed AI to speed up existing tasks – think automated QA, document synthesis or replace deterministic agents,” he expanded. “However, we’re seeing a sophisticated shift: companies are no longer just ‘bolting on’ AI, they’re redesigning processes to align with what agents can now do.”
AI has accelerated into the centre of business and consumer life but its meteoric rise has exposed deep and unresolved issues in relation to data readiness, governance, integration and ownership.
“Governance gets a lot of airtime but in our experience it’s usually solvable; alignment is achievable, even if slow,” challenged Patel of Mantel. “The harder problem is integration and the dependencies across multiple system surfaces, which requires careful management of business requirements, product leadership and prioritisation.”
Underneath that, Patel advised that the “deeper risk” is connecting the dots between the technology and the business. Product owners don’t always have visibility into where the technology is heading and what it’s capable of, and engineering teams who know agents don’t always have the context to reimagine a business process from the ground up.
“Bridging that gap is where a lot of the real work happens,” Patel added.
“When an operating model involves placing the business team directly with the engineering teams for agentic solutions, problems are solved with greater speed.
“Technical feasibility can easily be gauged against the true business value of solving a given problem agentically, leading to better prioritisation. It also follows that the business will ultimately be more comfortable signing off on the risk if they are directly involved in designing the human-in-the-loop mechanisms that ensure the evaluation of a solution.”
For Murphy of Onigroup, Google Cloud has done a “phenomenal job” of lowering the technical barriers.
Tools such as BigQuery and Gemini Enterprise Agent Platform have simplified data readiness and governance to a point where the ‘tech’ is no longer the primary bottleneck.
“The real friction is change management and psychological safety,” Murphy identified.
“There is a ‘hangover’ from the early days of GenAI where hallucinations were a valid concern. We spend a lot of time educating clients that those were the growing pains of early, unconstrained models.”
Today, through grounding, models can be wrapped in enterprise-grade guardrails. The market has moved from ‘experimental’ AI to ‘governed’ AI, where the frequency of these issues has plummeted in favour of high-fidelity, task-oriented precision.
“For example, a year ago, one of our junior engineers piloted a tool to automate our development QA testing,” Murphy explained.
“After his demo, he looked at me and said, ‘if this is successful, I won’t have a job.’ I had to stop him right there. I told him, ‘by automating the mundane, you’ve just become one of the most important architects in this business.’ He is now our biggest AI advocate.
“That fear is prevalent across the C-suite and the workforce. Without clear ownership at the business level and a culture that supports AI-driven evolution, projects stall. We find that the most successful organisations aren’t just ‘buying AI’; they’re actively rebranding it as a ‘capability multiplier’ for their people.”
Based on market experience – and addressing the technical challenges head-on – King of Aviato highlighted foundational data security, integration complexity and the misconception of data readiness as the biggest hurdles to scaling agentic AI.
“Google Cloud has tools for most of this, and governance is easy with 10 agents, but as you get to 20 you need proven processes in place (the HR department for agents) because at scale you need that level of control,” King said.
“Data readiness discussions often end with a light bulb moment with our customers. We have customers who want perfectly curated data sources before they talk AI but we’re not optimistic about this mythical data warehouse where all data is perfect.
“We can use AI to find and fix data issues, however, making the data clean enough and continuously kept to a high standard, so that AI can find insights or act on it. Using AI to get your data ready for AI is a huge shortcut.”
Data privacy and security came out as major themes of Google Next ’26 – highlighted by the launch of the Agentic Data Cloud, which uses a Knowledge Catalog and cross cloud capabilities like Apache Iceberg to unify data without massive migrations, allowing the agents built to act securely.
“Retail customers are really looking at how they can implement the concierge like tools in websites that have a lot of technical debt from years of add-ons, and with e-commerce being a huge part of their businesses these projects carry a lot of risk,” King added.
In a direct response to the roadblocks halting agentic AI progress in Australia, Bernadou of NCS stressed that governance, security and integration concerns “aren’t unknowns” and therefore can be answered and architected.
“The Google Cloud ecosystem is mature,” he shared. “In less mature platforms, these controls are still emerging – that’s why Google Cloud is so good at supporting third-party providers with their TPU capabilities.
“However, speed is more an overall market problem than a technical one. The real bottleneck is ‘tacit knowledge’, you can’t automate a process that only sits in the minds of employees. For AI to deliver broad productivity gains it needs to ‘understand’ the business and much of this knowledge has never been captured before.”
In Australia, the conversation around agentic AI is often centred on technology but its success will ultimately depend on people, expertise and ecosystem collaboration. The organisations that capture the greatest value will be those able to combine technology, industry knowledge and execution at scale.
In that regard, the strength of the Google Cloud ecosystem is not measured by the number of partners it has, but by the impact those partners create. As agentic AI moves from promise to production, ecosystems built on capability, co-innovation and customer outcomes will become one of the most important competitive advantages in the market.
Inform your opinion with executive guidance, in-depth analysis and business commentary.
