DataSapien targets AI ROI crisis with device-native marketplace

DataSapien targets AI ROI crisis with device-native marketplace

London-based DataSapien launches open beta for its Device-Native AI platform. The marketplace shifts intelligence from the cloud to local devices, aiming to address a $109 billion shortfall in enterprise AI returns.


DataSapien has launched an open beta for its Device-Native AI (DNA) platform — a new marketplace that allows application publishers to deploy artificial intelligence directly on user devices rather than through the cloud.

The launch comes as enterprises confront what analysts are calling an AI “ROI crisis”. Global investment in artificial intelligence has topped $109 billion, yet up to 80% of initiatives have failed to deliver measurable value, according to multiple industry studies. The company’s localised approach positions it within a growing movement to make AI cheaper, faster, and more privacy-resilient.

Unlike traditional cloud models, which charge per interaction and rely on remote processing, DataSapien’s architecture enables apps to run smaller open-source models on-device. The company says this removes “per-token” usage fees and cuts exposure to external data centres.

Co-founder StJohn Deakins, a technology veteran best known for coining the term omnichannel, said the aim was to return control to developers and users. “Businesses are currently trapped,” he said. “To get smart AI, they have to hand over their customer data to Big Tech and pay a ‘tax’ on every interaction. It’s expensive, slow, and dangerous.”

DataSapien’s marketplace currently supports models from Google, Meta, Microsoft, and MIT spin-out LiquidAI. Early beta testers have reported higher engagement rates and lower operational costs when compared with cloud-based services, though independent benchmarks have yet to be published.

The company’s DNA platform uses what it calls Zero-Shared Data architecture, allowing personalisation without transmitting private information off the device. That model aligns with tightening data-sovereignty and privacy regulations in both the UK and EU, which are pushing developers to limit cross-border data flows.

Analysts say the post-cloud model reflects a broader correction in the AI sector as investors reassess business cases built on heavy infrastructure costs. Edge-based approaches are attracting attention from telecoms and device manufacturers looking to reduce latency and improve user trust.

DataSapien’s open beta gives app publishers access to a no-code design interface and a software development kit for integrating models. More information and performance benchmarks are available at datasapien.com.



  • When AI stops advising and starts acting

    When AI stops advising and starts acting

    AI is moving from assistance towards delegated action inside chat. Tencent’s latest WeChat move points to a wider shift in enterprise technology, where the real question is no longer whether employees use AI, but how companies govern permissions, approvals, audit trails, and accountability once software begins acting on a worker’s…


  • ICS.AI targets university AI access gap

    ICS.AI targets university AI access gap

    ICS.AI is offering universities wider governed student AI access nationwide. The company says the model removes a major cost barrier and extends enterprise-grade access once institutions deploy its staff platform.


  • Meta breach exposes agent oversight gaps

    Meta breach exposes agent oversight gaps

    Meta incident spotlights fresh risks from autonomous workplace AI tools. RAIDS AI says the episode shows how trust in agent output can become a security weakness even without privileged system access.