Treasure Data Brings AI Marketing Cloud and ‘Super Agent’ Capabilities to AWS Marketplace

When I recently spoke to Kaz Ohta, I wasn’t expecting a lesson in reptilian neurology. But he was making a scaly point – and doing it in his own brilliant way.

“Back in 2000, when I was a student, I was working on supercomputers,” he said. “Around that time, my professor said, ‘Hey guys, I'm building an alligator.’ And I was like, alligator?"

As Kaz explained, the number of transistors in a computer processor 25 years ago was the same as the number of neurons in an alligator’s brain – in the tens of millions. People, on the other hand, have around 86 billion, and that's a big leap. Now, modern computers are approaching a comparable scale and a similar evolution in intelligence.

“That's why AI is really interesting these days,” Kaz said. And to get more targeted, it's also analogous to the growth beyond GPT-2 – which, when originally released in 2019, was only capable of handling two-second coding tasks. Today, with the exponential improvement of models, we’re seeing one-hour jobs. 

In other words, see ya later, alligator. AI is evolving to a whole new genus in the computing kingdom.

Treasure Data Co-founder and CEO Kaz Ohta. Source: LinkedIn

This maturation in model performance has been described as a new Moore’s Law for AI agents, which should equate to transformational performance. But as Kaz explained, the problem is the failure rate of these lengthy hauls, which hovers around 50%. 

That's predicted to improve over time, but today, it's a flip of the coin on success. And that's not a bet most businesses are willing to bank on. That said, right now, smaller agent tasks are delivering solid outcomes, with a closer to a 99.9% success rate. So what if there was a better way to orchestrate a myriad of smaller tasks while reducing the overall complexity?

Source: Treasure Data

That’s where Kaz sees an opportunity to deliver real value with Treasure Data – the company he started back in 2011. A leading force in the pure play customer data platform (CDP) landscape, I’ve been watching them closely as they’ve evolved their posture through the advent of agentic AI.

“We're obviously a pure play CDP player, and we focus solely on enterprise, so a majority of our customers are global brands with a billion-plus in revenue,” he said. “We have about eight billion people on earth, and five billion people are internet-connected, and then we have about two to three billion profiled inside our system. Obviously, we're not owning their data. But you know, we have an influence on billions of consumers and people, and at the same time, we have to be a custodian of that data.”

Making that data more valuable is the key – and AI is unlocking it. Back in October, at the company’s signature CDP World event, I wrote about the release of Treasure Data’s AI Marketing Cloud. More than just a product, it represents a sort of tipping point for the company, a transformation from CDP pioneer to an AI-native marketing cloud built on a single, trusted data foundation to unify intelligence, cross-channel orchestration, and real-time personalization.

Last week at the colossal AWS re:Invent conference in Las Vegas, Treasure Data announced that its AI Marketing Cloud is now available in the AWS Marketplace, giving global brands and companies faster, more direct access to its AI-native marketing platform. Traditionally a DevOps store for cloud-related tools, AWS Marketplace has evolved into a business application hub that harnesses the capabilities of Amazon’s best-of-breed cloud services.

Unlike stitched-together clouds, AI Marketing Cloud is AI-native, embedded from the ground up with Treasure Data’s Intelligent CDP – and not bolted on as an afterthought.

AI Marketing Cloud enables marketers and brands to deliver hyper-personalized experiences across every channel. By unifying trusted customer data, secure AI agents, and intelligent activation – and making it all available through AWS – enterprises can gain the efficiencies of lower TCO, mitigate dependencies on fragmented marketing and data tools, and power revenue growth through more relevant engagement and enhanced AI decisioning.

Composability is key, and while AI Marketing Cloud is a fully unified solution, it’s interoperable in all the right dimensions. Enterprises can extend it with their preferred tools and ecosystems, connecting seamlessly with data warehouses, analytics platforms, and activation partners. Armed with this hybrid flexibility, customers are empowered with choice and control, avoiding lock-in while gaining the power of a unified, intelligent data foundation.

Along with AI Marketing Cloud, customers can also procure Treasure Data AI Agents and its industry-leading Treasure Data Intelligent CDP from AWS Marketplace with ease – enabling marketing, data, and IT teams to discover, buy, and deploy its AI-enabled marketing platform through the hyperscaler while benefiting from the security, scalability, and simplified billing that AWS provides.

Delivering next-level orchestration with the Marketing Super Agent

As Kaz told me, there’s already been an incredible response to AI Marketing Cloud from Treasure Data’s customers, which enables features like email, mobile/SMS push, and more. 

“CDP was separated from those engagement layers, but now we have it,” he said. “But really what we want is to have AI orchestrating the complete AI journey.”

That’s where Treasure Data’s Marketing Super Agent comes in. Originally unveiled at CDP World in October, Marketing Super Agent is a powerful AI marketing resource built into Treasure Data’s AI Marketing Cloud that maintains its stack-agnostic flexibility.

Unlike traditional copilots that respond to prompts, Marketing Super Agent plans, creates, and executes marketing work end-to-end by orchestrating specialized task agents, maintaining persistent memory, and operating as the hub of a marketer’s daily workflow.

Each agent collaborates through a shared live memory layer, enabling true cross-workflow reasoning and delivering an AI that understands context, builds reasoning chains, and executes multi-step marketing work with precision. It dynamically assembles and coordinates specialist task agents, including deep research, sentiment analysis, persona development, competitor intelligence, campaign concepting, channel-specific ad generation, and more.

Treasure Data's Marketing Super Agent orchestrates smaller agent functions. Source: Treasure Data

During a demo of the Super Agent, Kaz walked me through the highlights, illustrating how it commands an army of sub-agents to achieve specific goals. Again, as he explained, these smaller tasks are highly successful in delivering outcomes for marketers.

“What you're going to have is this one agent that's orchestrating,” he shared. “But then underneath, we're adding a bunch of smaller agents.” 

In his example, he quickly prompted research on yours truly, bringing up general details about me, CMS Critic, and the Critic’s Corner podcast. He was then able to quickly summarize everything into a visual PowerPoint. From there, the orchestrator agents were able to create a plan in which other small agents could take over.

The Workspace UI itself is clean, intuitive, and easy to navigate. Everything feels like it's within arm's reach, and a further testament to Treasure Data's user experience. Once your plan is in place, it's simple to “Accept & Execute” the plan and move into next steps in the proper order.

Source: Treasure Data

One thing Kaz pointed out is where human checkpoints can be added (the essential “human in the loop”). As he said, this further enhances the results.

“We think if you're adding a lot of agentic workflow, the success rate goes down,” he said. “But if you have a human checking in the middle, the success rate of this task goes up really well.”

You can also change your plan at any point, or skip the orchestrator agent and go directly to a specific agent in Treasure Data’s Agent Hub to perform tasks like data analysis. From here, you can produce sweet visualizations like graphs and pie charts.

And that’s just scratching the surface. You can then leverage resources like Treasure Data’s Journey Optimizer Agent to recommend and implement customer journey changes and A/B tests to deliver personalized customer experiences. There are numerous enterprise-grade agents available in its wheelhouse that are campaign-ready.  

One of the most intriguing examples Kaz shared was the Personas Agent, which creates AI-driven personas for “role playing.” Users can interact with these personas via an Agent UI, which showcases details and allows journey creation for specific campaigns. It was both fun and enlightening to see this working with a fictional customer while crash-testing some real-world scenarios. 

“This takes random data from CDP and injects the customer data into this agent, and then all of a sudden, this persona role plays and acts as if its this customer,” he said. “You can actually use this virtual persona to come up with and test your campaign. So this orchestrator agent is now brokering the conversation between me and this virtual agent.”

From a composable perspective, Kaz is focused on extending the value of the AI Marketing Super Agent and making its capabilities extensible to other platforms. 

“I'm also here to open up our system too, so any DXP or outside vendor can talk to our Marketing Super Agent,” he explained. “We’re also adding support for MCP, so as long as the vendor supports MCP, our Marketing Super Agent can talk with their system.”

Trade-Up program and AWS Marketplace pricing

As part of its announcement at AWS re:Invent, Treasure Data has expanded its Trade-Up program to help enterprises consolidate their growing morass of martech tools while reducing costs. 

If you replace your legacy platform with Treasure Data’s Engagement AI Suite, which is part of AI Marketing Cloud, you can get up to 24 months free with a multi-year commitment. There’s also a CDP upgrade, and if you switch to Treasure Data’s Intelligent CDP, you only pay in full once your existing CDP contract ends. You can contact their sales team for more details. 

One of the benefits of procuring through the AWS Marketplace is relative price transparency. I previously covered how Treasure Data moved to a “No Compute” pricing model, but its AWS Marketplace listing pages provide an even more simplified baseline for its hosted products. For example, a 12-month contract for AI Marketing Cloud has a list price of $90,000. 

As a long-time AWS cloud practitioner, I would recommend carefully assessing the additional usage and infrastructure costs. Understanding the cloud economics is key – and having DevOps professionals consult on your deployment is always a good idea when executing a cloud-first strategy for any enterprise software application.

The verdict

Treasure Data continues to evolve into a treasure trove of marketing automation. As I’ve covered in recent news, they’re addressing the friction around complexity and cost. Now, with expansion into AWS Marketplace, they’re reducing the hassles surrounding procurement, deployment, and management.

Again, as with any cloud deployment, I would strongly advise an expert hand or DevOps team that understands the mechanics – just to ensure that everything is working within predictable parameters. Treasure Data does offer a bench of resources, including CDP training, as well as a Value Calculator that can help build your CDP business case. So tap into them directly as you weigh your deployment options.

AI Marketing Cloud is presenting solid features and capabilities with its agent-forward posture. Other marketing clouds offer their own modicum of orchestration and personalized customer experiences, but they historically pack in complexity and a lack of seamless ecosystem interoperability. This latent fragmentation of disconnected data models, inconsistent customer profiles, and unpredictable costs has put pressure on performance. This is where Treasure Data is making its impact felt with an AI-native intelligent CDP at its foundation.

I see a lot of potential in Treasure Data’s model for its Marketing Super Agent and orchestrating smaller agents for a variety of tasks. Agent performance is riddled with challenges; according to Cleanlab, 70% of regulated enterprises have to rebuild their AI agent stack every three months or sooner. With this approach, they’re effectively minimizing the risk of failure in unstable systems while harnessing the best performance from smaller AI tasks. 

A big part of the story is also the consolidation of SaaS and simplifying the morass of martech software. As Kaz noted in our conversation, despite there being over 15,000 SaaS vendors in the martech landscape, we've seen an 18% decline of SaaS apps in the stack over the last two years. 

This drift towards unification and ease of use is due in part to AI-enabled efficiencies, but also attributed to cost reduction and mitigating the cognitive overload of managing too many panes of glass. Marketers need better tools and solutions – and this is where Treasure Data is focusing its efforts.

According to Teradata, agentic AI might be approaching its market moment in 2026. While only 9% of organizations have fully adopted some form of agentic AI, nearly three-quarters express moderate to high confidence in its potential, suggesting positive expectations. Given its product evolution and market messaging, Treasure Data is well-positioned to meet the opportunity with differentiated value.

Next year, in another example of its shift towards autonomous AI, Treasure Data has rebranded its popular CDP World event to Agentic World, which will be held on October 6-7 in Miami. The conference is dedicated to the future of AI-powered customer experience, bringing together marketing, CX, and data leaders to discuss best practices for leveraging AI agents to drive growth and efficiency.

As Kaz said, AI is really interesting these days. Despite the technological advancements and non-stop cycle of innovation, it’s still about enabling human marketers with more knowledge and automation – so we can activate our 86 billion neurons to work smarter, and not harder.

If you're ready to evolve, Treasure Data is worth a look. 

See ya later, alligator. 

Upcoming Events

CMS Kickoff 2026

January 13-14, 2026 – St. Petersburg, FL

Meet industry leaders at our fourth annual CMS Kickoff – the industry's premier global CMS event. Similar to a traditional kickoff, we reflect on recent trends and share stories from the frontlines. Additionally, we will delve into the current happenings and shed light on the future. Prepare for an unparalleled in-person CMS conference experience that will equip you to move things forward. This is an exclusive event – space is limited, so secure your tickets today.