Cloudera describes itself as the company that brings “AI to data anywhere” today. It’s a claim stems from its work that spans a multiplicity of data stacks in private datacentres, in public cloud and at the compute edge. As we now witness enterprises move quickly through what the company says are “new stages of AI maturity”, Cloudera used its Evolve25 flagship practitioner & partner conference this month in New York to explain the mechanics of its mission to help enterprises navigate this transformation with an AI-powered data lakehouse.
This year’s keynote address was opened by “fun Aussie emcee speaker” Paul Muller, who gave way to Brian Russo, SVP for Americas at Cloudera. The opener then moved onward to Cloudera CEO Charles Sansbury. Coming from a chief financial officer background and now directing the ship at Cloudera, Sansbury says that his go-to factors that help him decide whether to help lead a company are a) its ability to scale b) whether it has a defined tailwind [for growth, expansion & innovation] and c) it helps when an enterprise works with investors that he knows.
Speaking on stage in New York this year, Sansbury said that the company is now overseeing 25+ exabytes of data under its platform.
“We’re now an important part of the discussion when customers are analysing how they are going to really operationalise AI. Leaders across every major industry in data-driven organisations use Cloudera to power their AI projects… but things have changed i.e. AI has moved onwards from being used [just] for anomaly detection to really being used in business user environments that drive real business value,” he said.
Talking about how AI “reduces the human need” in tasks that can now be executed by an AI agent, Sansbury is upbeat about where this allows organisations to now re-task their human workforce on more intuitive decision-making in higher-value jobs, roles and functions.
Hybrid: the end-state architecture of choice
“The endpoint architecture for a business’s data estate is hybrid; it is emerging as the end-state architecture of choice for most progressive organisations. This means that companies need full operational control from the datacentre to the edge in terms of their application of data and how it is fused and channelled to the AI they want to create, deliver and use. We are investing in the future based on feedback from customers to continue to deliver a platform that enables firms to utilise data wherever it exists,” said Sansbury.
Taking over from the CEO, Frank O’Dowd, Cloudera chief revenue officer drilled into why he thinks the company is all about being able to enable customers to run the right data workload on the right platform. Given the escalating costs of cloud, the need for technical expertise and the legacy platforms that are out there (some of which aren’t too old i.e. legacy happens a lot faster these days, obviously), O’Dowd showcased just how many AI projects today are failing.
Vendors as ‘adaptive service partners’
“But what do the winners in AI do? They’re treating vendors as ‘adaptive service partners’ and working with them to solve narrow problems that can be validated and delivered upon in concrete terms, first and foremost. Our data lakehouse is behind what we do and we’re now providing data services with cloud-native tooling across the private cloud and the public cloud – we are doing this because we understand the investments that customers are making in hybrid infrastructures,” said O’Dowd. “But because there is a need for both public and private cloud now – more pressingly than ever – we also need to bring those two worlds [of technology] together now.”
O’Dowd also pointed to the Cloudera AI Inference service. This is a technology designed to accelerate the ability for AI models to serve, deploy and scale private AI applications, agents and assistants with speed and security. “Cloudera AI Inference lets [organisations] unlock data’s full potential at scale with Nvidia’s AI expertise and safeguard it with enterprise-grade security features so [users] can confidently protect data and run workloads on-premises or in the cloud, while deploying AI models efficiently with the necessary flexibility and governance, said Sanjeev Mohan, principal analyst, SanjMo.
What is an AI maturity curve?
Next up to deliver the CRO keynote, the audience welcomed Abhas Ricky in his role as chief AI product officer. Ricky talked about the development of real AI maturity curves today in business and said that we have moved through various stages this decade that can be defined as a move from a) AI curious b) AI enabled c) AI embedded to d) AI-native, where we actually start to see autonomous agents working inside operational businesses. He says that what enterprises really need to reach intelligent autonomy with governed AI agents is a knowledge hub, a fully autonomous scalable data platform and a governance framework to be able to drive the intelligent enterprise that every organisation now wants to become
“An AI-powered data lakehouse is a control tower,” said Ricky. “A seamless UI enables organisations to straddle public cloud on one side and public cloud on the other. This is all about grasping a platform that is capable of scaling autonomous decisions. We’re in the business of helping organisations get intelligence inside better, faster and cheaper, but the hardest thing to do (as a customer in Singapore told me last year) is getting high fidelity data ‘slash’ context to an organisation’s IT stack for AI today.”
Further sessions on product & strategy were delivered by Leo Brunnick in his role as chief product officer at Cloudera and, to really get into the technical guts of the company’s platform, Serio Gago Huerta, chief technology officer at Cloudera.
Cloudera used this event to announce the expansion of its Enterprise AI Ecosystem with new partnerships designed to deliver production-ready AI solutions. This initiative brings together industry leaders to offer a suite of solutions that address enterprises’ most pressing end-to-end AI needs.
AI now moves to ‘advanced territory’
As we know, organisations have moved through focusing on stages of AI maturity from retrieval-augmented generation (RAG) to fine-tuning and copilots. Today, Cloudera thinks that adoption has “accelerated into advanced territory”, by which it means predictive engines for structured data, AI-driven workflow automation, observability for model reliability and large-scale document intelligence. In this new realm, we can see that companies are embracing new model choices, orchestration frameworks and agent-based systems capable of complex reasoning and multi-step task execution.
“With this approach, organisations can move beyond experimentation to embed AI directly into business operations, unlocking high-value use cases across customer experience, fraud detection, supply chain forecasting, IT operations and compliance, all while maintaining governance, security and architectural flexibility,” noted Sansbury and team, during this year’s Evolve25 event. “Cloudera’s end-to-end platform delivers both AI-ready data and AI agents that help to transform that data into intelligent actions. These tools have already enabled hundreds of Cloudera customers to shift from AI experimentation to having AI embedded throughout their functional teams.”
The company’s ultimate goal is to help its customers become truly AI-native and be able to use autonomous decision-making with minimal human intervention, supported by trusted, transparent systems. To accelerate this vision, Cloudera is expanding its Enterprise AI Ecosystem with four new partnerships.
- ServiceNow: ServiceNow’s enterprise workflow automation and AI-powered products will now see a future integration that will feature ServiceNow’s Workflow Data Fabric zero copy connector with Cloudera’s data foundation to allow organisations to securely access real-time enterprise data without duplication across IT, HR, finance, customer service, compliance etc.
- Fundamental: Fundamental provides a prediction engine for enterprise tabular data. Most enterprise challenges, including churn prediction, credit risk, fraud detection and demand forecasting, are tabular prediction problems that have yet to be impacted by deep learning that have yet to be impacted by deep learning. Fundamental addresses this gap with a foundation model that requires no parameter tuning or feature engineering.
- Pulse: Pulse provides a document processing engine, turning unstructured content (e.g. contracts, claims reports etc.) into structured, LLM-ready data. By integrating Pulse into Cloudera’s AI-powered lakehouse, enterprises can automate data flows from document ingestion directly into ERP, CRM and compliance systems.
- Galileo.ai: Galileo specialises in AI observability, helping enterprises validate, monitor and maintain their AI systems in production. Its platform tracks model accuracy, drift, and reliability in real time, with dashboards and alerts purpose-built for large language models and agent-based systems.
With Cloudera’s added capabilities, Galileo provides a “closed loop for trusted AI deployment” say the firms. Data flows into the Cloudera lakehouse, models are trained and run on that data and Galileo provides the visibility to ensure those models remain accurate, fair and reliable as conditions change. Whether monitoring predictions generated by Fundamental’s tabular foundation model or insights extracted from Pulse’s document pipelines, Galileo ensures that every AI-driven workflow built on Cloudera stays compliant, transparent, and high-performing. This combination allows enterprises to not only deploy AI at scale, but to do so with the confidence that outcomes remain trustworthy over time.
Operationalising AI & agentic workflows
“The Enterprise AI Ecosystem has become a cornerstone of our strategy to help large enterprises navigate the complexities of AI adoption,” said Cloudera CPO Ricky. “Our newest partners bring specialised capabilities that directly address the biggest challenges our customers face today: operationalising AI and agentic workflows at scale with ServiceNow, ensuring transparency, reliability, and accuracy with Galileo.ai and Pulse, and unlocking the next generation of AI on structured data with Fundamental.”
“Fundamental makes predictive AI on tabular data simple and powerful,” said Jeremy Fraenkel, CEO and Founder of Fundamental. “Unlike foundation models trained on text or images, ours is purpose-built for the structured data that runs every enterprise, from transactions to customer records. Partnering with Cloudera, enterprises can now apply this predictive foundation model across their most critical datasets without the complexity of custom pipelines or tuning.”
Existing members of Cloudera’s AI Ecosystem include capitalisation-focused GPU company Nvidia, AWS, Pinecone, Google Cloud, Anthropic, Snowflake and CrewAI. Cloudera explains that its partner network is rooted in its “foundational belief” that “no single vendor” can solve all the intricate requirements of large-scale AI deployment… so the company insists on fostering an open ecosystem.
Data estates are a mess, let’s clean up
Cloudera CEO Sansbury left us with some parting word in a private press session and said that “most company’s data estates are a mess” today… so this is clearly why his firm is so singularly bent (well, multilaterally-aligned for all hybrid cloud and device possibilities in fact) on getting those messy data estates into shape and creating fluid channels for all data to be able to flow into the right workflow, through the right intelligence engine and forward through the right AI service.
That could just be where intelligent intelligence goes next…
