Acceldata launches AI platform for hybrid data estates

Acceldata has launched its Autonomous Data & AI Platform, designed for organisations running data across hybrid and sovereign environments.

The launch follows research commissioned by Acceldata showing that nearly 80% of enterprises use hybrid data operations, while more than 40% cite governance fragmentation as their biggest challenge in cross-platform environments.

Acceldata is positioning the platform as an alternative to architectures centred on consolidating data into a single lakehouse. It argues that many businesses now operate across multiple systems and are under growing pressure to support artificial intelligence projects without undertaking lengthy migration programmes.

Research conducted by GLG for Acceldata surveyed C-level executives at Fortune 1000 and Global 2000 companies. It found that 80% of enterprises with revenue above USD $5 billion use data warehouse and data lakehouse architectures together, while 75% run four or more data platforms in production.

That mix has created operational strain. More than 40% of respondents identified governance fragmentation as the biggest challenge in cross-platform data environments, while 20% pointed to data duplication and 10% cited identity fragmentation and data lineage gaps.

Separate findings highlighted the persistence of older systems. Some 43% of enterprises said they still had at least some Hadoop workloads in place, while 20% were in the process of decommissioning workloads.

Regulation also remains a major factor in infrastructure decisions. The survey found that 57% of respondents cited regulatory requirements as the main reason for on-premises deployments, ahead of legacy inertia at 54%, with data sovereignty and technology debt both at 50%.

Hybrid pressure

The new platform is intended to let analytics workloads and AI agents run where enterprise data already resides, including in cloud, on-premises, hybrid and sovereign environments. It includes compute, governance, observability and data quality functions in a model Acceldata describes as xLake.

According to Acceldata, the platform routes workloads across different infrastructure locations, applies governance controls and supports autonomous operation across distributed data sources. It is also intended to give enterprise agents broader access to data across business processes.

Cost pressure is another part of the backdrop. The research found that support for AI initiatives is the leading source of board-level pressure on data infrastructure, cited by 33% of respondents. Companies also reported friction in AI and machine learning operationalisation, integration gaps and skills shortages.

Respondents also raised concerns about the economics of AI projects. Compute cost volatility, weak cost governance and unpredictability in usage-based charging were identified as recurring problems.

Acceldata argues that these issues reflect a mismatch between AI ambitions and the design of many existing data architectures. Enterprises, it said, are trying to run AI programmes on systems not built for hybrid operations, distributed governance or the cost controls now required for large-scale AI use.

Rohit Choudhary, Founder and Chief Executive Officer of Acceldata, said the shift away from centralised architectures had been building for years. "The lakehouse architecture was built for human access. It broke in the agentic era," he said.

He linked that shift to regulatory change in Europe and to the broader difficulty of consolidating enterprise data into a single environment. "We started Acceldata with the conviction that enterprise data would never consolidate, that hybrid would be the durable reality and that the data and AI platforms must evolve to support it. In Europe, data sovereignty mandates are accelerating this shift, making hybrid-native, jurisdiction-aware architectures a board-level imperative, not a future consideration," Choudhary said.

The launch adds to a broader push by infrastructure and data management suppliers to adapt products for customers with fragmented estates spanning public cloud, private systems and regulated local deployments. For large companies, the challenge is not only how to run AI models, but also how to govern access, control costs and maintain data quality across a patchwork of platforms.

In Acceldata's survey, that patchwork appeared to be the norm rather than the exception, with three-quarters of respondents saying they now manage at least four production data platforms.