“Agentic AI is a real game changer”: Australian partners react to Google Cloud Agentic AI updates
CRN Australia speaks to three partners on the new AI-based features from the search engine giant.
Google Cloud Next has come and gone, and with it a slew of new AI applications and partnerships announced, the talk of the town being Agentic AI.
During the event last month, Google Cloud CEO Thomas Kurian said, “Agentic AI is the biggest opportunity for partners, with billions of dollars of economic impact in terms of productivity from delivering agents within customers. That’s where we see the future.”
Closer to home, Gary Denman, head of partnerships and Alliances and Google Cloud ANZ told CRN Australia that customers and partners are looking to the search engine giant for Agentic AI capabilities.
“What we're hearing is, number one, the results and what they're trying to achieve is being achieved at a great quality, a better quality than other solutions.
“Number two is that they're able to move faster and number three is that it's coming in at a cost performance basis, which is what we need.”
But what do Australian partners think about Agentic AI?
CRN Australia spoke to three partners about Google’s new announcements and what it means to them as Agentic AI begins to be front of mind for partners and customers alike.
“We’re getting tonnes of work from building AI agents”
Ben King, founder at Aviato Consulting attended Google Cloud Next in the US, noted that one of the key things he likes about Google is the cost of the AI models are so much cheaper to run.
“You can run Anthropic on Google, and it's about 100 times the cost when you're using the API. I was like how's Google doing this? Are they just losing money on it?
“But they released their new TPU, the seventh generation one, and it's called Ironwood. It's so much cheaper for them to use those rather than Nvidia that everyone else has to use.”
From the low cost and performance of this TPU, King is astounded by the amount of work that is being commissioned by his customers for Agentic AI applications.
“We're getting tons of work from building AI agents,” he said.
“Google's agent space was one of the other things they released a lot of connectors for. We can set it up in two hours. The hardest part is making sure that we've got the admin access to SharePoint or Atlassian’s Jira.
“The agents of people building for specific use cases based on using Gemini at the cost, it is mind blowingly fast, super cheap, and it's getting to almost be as good as humans doing that job.”
Aviato Consulting has been using Agentic AI since last year where they tested the warters with several proof of concepts.
King explained that this year they’ve seen an increase in demand from their customers for using Agentic AI, so much so that they are reportedly beating GSIs for jobs.
“We're doing three two week sprints and we'll have something running that we can show [customers] and then iterate on it from there,” he said.
"We're winning against big global systems integrators, just from being more agile. Those guys are still always going to write whitepapers and come up with the AI strategy, but for the implementation, they're just a bit too slow, which is awesome for us.”
More practical applications of AI
With all the Agentic announcements, Simon Poulton, partner, Google Cloud at Mantel Group said the practical side of Agentic is getting both him and his customers excited.
“Like the AI expansion and explosions...In terms of the practical nature of it, I think it's really moving forward,” he said.
The announcements made at Google Cloud Next have ramped up the application and deployment of these systems, Poulton said.
“There's a lot of very practical things that show that these systems are now getting deployed quite seriously by customers and it’s all quite exciting,” he noted.
Poulton added that he is excited about introducing the agent development kit and the agent-to-agent protocol from Google.
While all these practical applications are causing hype at Mantel Group, the issue of safety around Agentic AI has been brought up.
“The question for most organisations, I think they're facing is, how do we [implement Agentic AI] that safely?” Poulton said.
“How do we do it in a way that will get the return on our investment and avoid disruption?”
He continued, “That's another important consideration, because when you’re going to produce something substantial as an organisation, all of those concerns around doing it safely, avoiding reputational risk and making sure there's traceability and explainability, those things become critically important functions.
“Most executives I speak to that's very high on the radar.”
“Agentic AI is a real game changer”
Looking back at the conference, Joel Collins, founder and director at data consultancy intelia said the clear acceleration and focus on Agentic AI solutions was the major standout.
“Globally, businesses are not only experimenting but fundamentally reimaging the way work gets done,” he explained.
“From Google-like search across the enterprise, to uncovering trends and insights in hidden data to transforming the way organisations engage with their customers, Agentic AI is uplifting organisational capability rapidly – and those who aren’t on the journey are being left behind.”
Collins said the formalisation and enhancement of tools for building AI agents, particularly within Agentspace, and the introduction of concepts like the Agent2Agent communication protocol, signal a “significant step” towards creating sophisticated, interconnected AI systems.
“These align perfectly with where we see the market focusing and where we are building our expertise,” he added.
Collins had nothing but good things to say when it came to Agentic AI and its capabilities.
“We think Agentic AI is a real game changer,” he said. “Not only in the capability it currently offers, but also in the way humans engage with software more broadly.
“Although it’s early days, we believe that the entire paradigm of human work will shift from AI being at the edges, to it being embedded into every step of the workday – from automating mundane tasks to deep understanding of problem statements and being able to autonomously solve issues and provide proactive advice.”
Collins said this means agentic capabilities will increasingly become a core component integrated within many organisation’s data, analytics, and infrastructure offerings, rather than just a standalone product.
“We think it’s a big few years ahead for technology providers and consumers,” he added.
In terms of how these new updates from Google will make them money, Collins said as primarily a services business by both directly and partnering with Google, they are taking sales plays to market focussed on Agentic solutions.
He highlighted two key areas of focus.
“Creating significant interest in new solutions that didn’t exist a few years ago, GenAI, Agentic Automation, resulting in new use cases being delivered in customer environments and increased interest in adopting Google technology,” he said.
“This directly drives demand for our consulting and implementation services.”
The second area is supercharging his team’s capability.
“We can deliver much more complex solutions at a significantly faster pace with all these new tools at our disposal, increasing our efficiency and capacity to take on more projects,” he ended.