San Francisco - After several years of work on Apple Vision Pro-related spatial computing initiatives, former Apple Design Engineering Lead Victor Hu (Yunqi Hu) recently founded Noscen, a San Francisco-based spatial intelligence company focused on developing human–AI interaction systems that integrate artificial intelligence, spatial computing, hardware, and real-time environmental context.
Hu founded Noscen in 2026 after several years of work on Apple Vision Pro-related spatial computing initiatives, where he contributed to next-generation human–AI interaction, generative AI features for future hardware platforms, and visionOS experience prototypes. At Noscen, his work focuses on translating those system-level interaction concepts into startup-led spatial intelligence platforms.
As spatial computing moves from experimental concept to real-world deployment, Hu’s work reflects a broader shift in computing: interfaces are no longer limited to screens, and AI systems are beginning to operate within physical environments, respond to spatial context, and assist users through continuous real-time interaction.
From Prototyping to Defining Interaction Paradigms
At Apple’s Vision Product Group, Hu’s work operated at the system level. Rather than focusing only on visual interfaces, he helped explore how generative AI, spatial computing, real-time interaction, and device-level constraints could be brought together in future user experiences.
His personal statement describes his Apple work as involving the development of AR and VR technologies for Apple Vision Pro, including the exploration, prototyping, and definition of use cases and human-computer interaction paradigms across immersive entertainment, personal productivity, remote collaboration, manufacturing training, and simulation.
“The real question isn’t just what AI can generate,” Hu said. “It’s how AI participates in your environment—how it responds, adapts, and collaborates with you in real time.”
This distinction is central to Hu’s work. In spatial computing, AI cannot function merely as a chatbot or a content generator. It must interact with the user’s environment, respond to spatial context, and operate within the constraints of real-time perception and rendering.
Integrating AI Into Real-Time Spatial Systems
A defining aspect of Hu’s Apple work was integrating AI-driven capabilities into performance-critical spatial environments. His résumé states that he spearheaded the conception and rapid prototyping of next-generation human–AI interaction across the Apple ecosystem and led multidisciplinary teams spanning research, design, machine learning, and software engineering.
This work required bridging high-level model behavior with low-level system constraints. In spatial computing, perception, rendering, gesture recognition, spatial mapping, and inference must operate simultaneously. If latency, power consumption, or rendering performance fails, the experience can break immediately.
Hu developed interactive software prototypes and production-quality features using Swift, C++, and Metal, leveraging ARKit and RealityKit on Apple Vision Pro. He also worked on 3D rendering pipelines, gesture recognition, and real-time spatial mapping, ensuring smooth performance under device constraints.
This combination of technical execution and interaction design differentiates Hu’s role from conventional product design. His work connected concept development, software prototyping, system validation, and user experience into a single process.
“In spatial computing, everything happens at once—perception, rendering, interaction, inference,” Hu said. “If any part breaks, the experience breaks.”
Enabling Complex and High-Stakes Use Cases
Although public discussion of spatial computing often focuses on consumer entertainment, Hu’s Apple work extended into professional and high-stakes environments. His résumé states that he helped define and prototype visionOS experiences including Persona SharePlay, surgical applications, and other unreleased features.
His personal statement further identifies enterprise-oriented use cases such as customized workspaces, collaborative 3D design, specialized training programs, and augmented support for remote fieldwork. These projects aimed to improve efficiency, productivity, and safety across different professional environments.
Such use cases require a different standard of interaction design. In surgical, training, or remote fieldwork contexts, spatial information must be precise, timely, and cognitively manageable. The interface cannot overwhelm the user, and AI-driven assistance must be integrated into the workflow without distracting from the physical task.
Hu’s work addressed these constraints by combining user research, rapid prototyping, and simulation-based validation. His approach focused not only on what spatial computing could display, but on how users could safely and effectively act on that information.
Leading Across Disciplines at Apple
Hu’s role at Apple also involved leadership across disciplines. His résumé states that he led a multidisciplinary team spanning research, design, machine learning, and software engineering to define and prototype visionOS experiences. He also collaborated with cross-functional teams of designers, artists, and engineers to translate product concepts into scalable, maintainable software solutions.
This leadership was especially important because human–AI interaction in spatial computing does not belong to a single discipline. It requires coordination between model behavior, product strategy, visual and interaction design, systems engineering, rendering, and hardware constraints.
Hu’s work frequently began at the “0 to 1” stage: defining the interaction model, validating feasibility through rapid prototypes, and iterating toward scalable implementation. His ability to move between vision, code, system design, and user experience positioned him as a key contributor to next-generation platform exploration.
Patent Activity in Extended Reality
Hu’s contributions at Apple also extended into invention disclosures and patent filings. Apple-related documentation provided for this matter identifies him as an inventor on multiple unpublished patent applications assigned to Apple Inc., with subject matter relating to virtual reality and extended reality technologies.
The available patent-related materials indicate topics including training and simulation involving spatial media content, and user interfaces for extended reality experiences incorporating live feeds from external sensors. Because these materials concern unpublished patent applications, they should be described carefully as patent applications, not issued patents.
This patent activity is significant because it shows that Hu’s work was not limited to internal prototyping or execution. It also contributed to protectable technical concepts in Apple’s extended reality and spatial computing research pipeline.
From Apple to Noscen
In 2026, Hu founded Noscen, where he serves as Founder and CEO. The company focuses on next-generation spatial intelligence systems that integrate AI, hardware, and spatial interaction. According to his résumé, Hu built Noscen’s software and hardware prototypes from zero to one and raised more than $5 million in funding.
This transition from Apple to Noscen reflects a broader trajectory in Hu’s career: taking complex interaction models from large-scale corporate research environments and translating them into startup-driven systems.
“Startups give you the ability to challenge assumptions at the system level,” Hu said. “Especially in a field where the interaction model itself is still being defined.”
At Noscen, Hu’s work builds on his prior experience at Apple and Niantic while moving toward a broader vision of spatial intelligence infrastructure. Rather than treating AI, hardware, and AR as separate technologies, Noscen is developing systems in which they function as a unified interface layer.
A Broader Impact on the Future of Computing
Hu’s work sits at a critical intersection in the evolution of computing: where artificial intelligence becomes spatial, embodied, and continuously interactive. His contributions at Apple helped explore how generative AI could operate within real-time spatial environments, while his work at Noscen extends that vision into a new company focused on spatial intelligence systems.
As spatial computing expands into enterprise, healthcare, education, remote collaboration, and simulation, the importance of system-level interaction design will only increase. The challenge is no longer simply whether hardware can display immersive content. The challenge is whether intelligent systems can understand context, assist users safely, and operate fluidly within the physical world.
For Hu, the goal remains clear.
“We’re not just building new tools,” he said. “We’re shaping how humans and intelligent systems coexist.”
Media Contact
Company Name: noscen, inc.
Contact Person: Yunqi Hu
Email: Send Email
City: San Francisco
Country: United States
Website: noscen.ai
