The next wave of consumer devices won’t capture the world; they’ll understand it. Spatially aware AR glasses, always-on computer vision, and persistent AI features all depend on something most hardware still can’t deliver: real-time perception within a tiny power budget. Today, those capabilities are largely confined to systems that can afford power-hungry application processors and often GPUs, putting truly wearable form factors out of reach. Mosaic SoC, which builds dedicated perception chips that bring spatial intelligence to energy-constrained devices was built to change that.
Today, the company announced a $3.8 million pre-seed round led by Founderful with participation from Kick Foundation.
Devices are gaining cameras and sensors faster than they’re gaining the intelligence to use them. The compute needed to interpret those signals still sits behind heavy processing stacks that drain batteries and force compromises in size, heat, and industrial design. For Original Design Manufacturers (ODMs) building next-generation AR and mobile hardware, adding more compute often means adding more complexity. Mosaic SoC takes a different approach: a dedicated perception chip that provides a baseline layer of spatial intelligence, with a full application layer that ODMs can integrate and build on top of.
Mosaic SoC builds integrated circuits that process visual and positional sensor data to give devices a real-time understanding of where they are and what’s around them. The company describes it as turning space into signals. The Mosaic SoC chips are designed to be small enough and efficient enough to make smart glasses indistinguishable from regular glasses, while still delivering full spatial awareness. The goal is to unlock device form factors that until now simply weren’t viable.
The chip lets a device build a local map of its surroundings and the objects within them, enabling features like recalling where an item was last seen or generating a floorplan on the fly. In smartphones, Mosaic SoC can act as a co-processor for the front camera, running always-on tracking and classification at a fraction of the power. That means a device can trigger recording only when a specific event occurs or a certain object appears, delivering continuous awareness without draining the battery.
The company was founded by duo Moritz Scherer and Alfio Di Mauro, both PhDs from ETH Zurich with deep expertise in system-on-chip architecture. They identified a widening gap between demand for edge intelligence and what existing hardware could actually deliver. The business model is straightforward: the company sells integrated circuits. But what makes Mosaic SoC unusual is that adding its chip doesn’t add complexity for ODMs. It removes it. The chip ships with a full application layer that Mosaic SoC develops and maintains, so ODMs can integrate it and build on top of it rather than engineering perception capabilities from scratch. The ambition is to bring spatial intelligence to every device where it was previously impractical.
In its first year, Mosaic SoC has already generated meaningful revenue through NRE contracts with ODM partners. As its chips reach the market, the company expects its revenue profile to shift from engineering engagements toward scalable product revenues tied to chip sales.
Mosaic SoC’s core differentiation is architectural. Where competing approaches rely on single- or dual-core ARM-based designs, Mosaic SoC uses a proprietary multi-core architecture with eight or more cores, engineered to maximize performance per watt and make always-on perception viable in energy-constrained devices. But the company sees hardware as just the starting point.
Mosaic SoC is building AI deployment toolchains and compilers that let firmware developers fully leverage the architecture, with plans to evolve from a chip provider into a platform supplier where applications are developed, deployed, and optimized around its silicon.
Looking ahead, Mosaic SoC’s goal is to become the standard layer for spatial intelligence at the edge, enabling always-on perception in wearables and mobile devices without the power and complexity tradeoffs that have held the category back.
Sage and PwC commit to tackling AI trust gap in finance
Posted in Commentary with tags Sage on April 30, 2026 by itnerdSage today announced a new initiative in partnership with PwC, which will redefine how AI is built and adopted in finance, combining transparent, explainable AI with the governance and real-world expertise required to use it with confidence.
The initiative, “Beyond the Black Box”, was announced at Sage Future, and is backed by new research from Sage, conducted by IDC, showing that more than seventy percent of finance leaders (71%) would reject an AI system if it cannot explain its outputs, even if they are highly accurate, showing that trust, not technology, is holding back AI adoption.
Unlike previous AI initiatives that have focused on large enterprises or purely technical audiences, “Beyond the Black Box” was created with SMB realities at its core. It forms part of Sage’s commitment to helping more SMBs benefit from the transformative impact of AI, building upon the company’s Responsible AI framework and AI Trust Label, reinforcing the belief that trust must be built into AI from the outset.
Trust, not technology capability, is the biggest barrier to AI adoption in finance
As AI becomes more capable, the ability to explain and stand behind its outputs is emerging as the defining factor in whether it is trusted and adopted in finance.
The consequences are already measurable. Finance professionals are spending an average of 12.9 hours every week reconstructing, validating and defending AI outputs. Much of this work stems from the need to validate and explain outputs that do not clearly show how they were produced. Rather than removing overhead, opaque AI is creating a new category of it.
Sage describes this as the trust cost of AI – the gap between what AI systems promise in theory and what finance teams can actually rely on in practice. At its core, this is a transparency challenge. Every number, recommendation and AI-supported decision must be explainable to auditors, to boards, and to regulators. When it cannot be, adoption stalls.
From black box AI to glass box
Sage has designed its AI from the ground up for the realities of finance, where every output is transparent, explainable and accountable, so organizations can trust and act on it with confidence.
This represents a deliberate shift away from black box AI, where outputs are generated without visibility into how decisions are made, towards what Sage describes as glass box AI: customers can meaningfully interact with AI results – not blind faith. Every answer is explainable, every recommendation is verifiable, and every output can be interrogated.
Through the initiative, Sage and PwC will combine their expertise into practical tools and frameworks to help finance teams understand, assess, and adopt AI responsibly. This includes embedding trust into how AI is implemented in finance environments while building on Sage’s existing commitment to SMBs, including the Sage AI Academy, which supports organizations with the knowledge and guidance needed to adopt AI with confidence.
From pilot to practice
To help move organizations from AI experimentation to trusted, scalable adoption, Sage selected PwC as its lead partner, drawn by PwC’s proven expertise in deploying AI across its own business. PwC has embedded AI into day-to-day workflows at scale, with 86% of its employees actively using AI tools, more than 240,000 Microsoft Copilot licences deployed, and over 4,000 custom GPTs developed and reused across the firm.
Businesses are increasingly concerned about the probabilistic nature of AI systems, particularly the lack of transparency, explainability, and clear accountability behind AI-generated outputs. Together, Sage and PwC will build transparent AI that gives finance teams control and full visibility into its outputs, backed by the implementation expertise, governance frameworks, and risk management capabilities required to put that AI to work safely, effectively, and at scale.
Leave a comment »