Big data lakehouse company Dremio Corp. today announced a new, highly automated and fully-managed version of its platform called Dremio Cloud, promising to free data engineers from drudgery and make them 10 times as productive as they were before.

Dremio is the creator of a specialized “agentic” data lakehouse platform that combines the capabilities of data lakes and data warehouses into a single repository to support autonomous artificial intelligence systems. It’s said to support fast data access and high-performance querying while providing a unified view of both structured and unstructured data. Through this architecture, it streamlines the ability of AI agents to get the insights they need to automate work, without the limitations of traditional data siloes.

It’s a comprehensive data platform, but adoption has been held up by the need for skilled data engineers to set it up and maintain it. Like all data architectures, Dremio’s data lakehouse is a complex beast. It requires constant manual intervention to set up and keep it up and running, and that means companies have to spend weeks training their data engineers on the intricacies of its architecture first.

Dremio’s solution to these headaches is to use its own AI agents to automate this work. Although it’s a fully managed platform, Dremio Cloud relies heavily on agents to continuously learn and optimize its settings in real time, without the need for intervention by human data engineers, the company said.

Because the platform is targeted at AI agents, it provides everything needed to simplify data access for autonomous workloads. At the core of Dremio Cloud is Dremio’s Open Catalog, based on the Apache Polaris project, which makes it simple to catalog and govern data, while access to both structured and unstructured data sources is facilitated by Dremio’s Intelligent Query Engine. The platform also includes an AI semantic layer, which acts as a kind of encyclopedia for business data, providing the context needed for AI to avoid hallucinations.

Dremio Cloud is also extremely accessible, with an agentic chat-based interface that allows business workers to query data and obtain insights without any knowledge of things like Structured Query Language. It also supports the open-source Model Context Protocol, allowing third-party agents powered by AI models from OpenAI Group PBC and Anthropic PBC to connect to it.

Agentic automation for agentic data

In terms of automation, Dremio Cloud introduces a new active metadata system that continuously optimizes user’s query patterns, data relationships and usage trends to support autonomous decision-making. The company describes this as a “living intelligence layer” that allows it to predict user’s needs and take preventive action to preclude any performance-related issues.

Elsewhere, Dremio Cloud’s enhanced semantics allow it to create wikis and labels automatically for things such as tables, views and more, so humans can easily access their data and check it for quality. Meanwhile, the automated clustering feature makes it possible for data layouts to be reorganized and restructured intelligently, in real time, to optimize query performance, the company said.

Dremio Chief Executive Sendur Sellakumar said companies can try the platform for free with $400 worth of credits and see for themselves how it fundamentally transforms the way data architectures operate. “By embedding intelligent, autonomous capabilities into a managed platform, Dremio Cloud eliminates the manual overheads that plague data teams, while providing the performance and reliability that AI agents demand,” he said. “As our agentic autonomous capabilities evolve, they will increasingly operate as coordinated agents themselves, creating a truly self-managing data ecosystem.”

Image: Dremio