Md Rafshan Jani@rafshan_jani721

Velos explores what becomes possible when intelligence moves from the cloud closer to the device. It is an AI-powered interaction workspace that combines computer vision, camera-based perception, gesture recognition, and real-time controls to transform a conventional computer into a more intelligent and context-aware interface. Rather than treating AI as a separate chatbot or cloud service, Velos explores AI as an interaction layer between the user and the device. Local perception can enable faster responses, reduce dependence on constant connectivity, and provide a more privacy-conscious approach to processing visual interactions. The architecture is designed with a pathway toward on-device AI inference, where computational workloads can be optimized across CPU, GPU, and NPU. This makes Velos particularly relevant to Qualcomm’s Model-to-Device Innovation vision and the Snapdragon ecosystem. Our goal is to demonstrate a future where devices do more than execute commands, they perceive, understand, and respond in real time. Sense locally. Understand locally. Respond locally.

🌎 Hybrid Hackathon | 💻 Online Build: September 10–16, 2026 | 📍 On-site: September 15–17, 2026 The official hackathon of AI Infra Summit — co-organized by Kisaco Research and lablab.ai. Build production-grade AI applications on the modern AI infrastructure stack, alongside 8,000+ attendees, researchers, founders, and infra leaders. 📍 On-site Venue (September 15–17): Santa Clara Convention Center 5001 Great America Pkwy, Santa Clara, CA 📍 On-site participation is by invitation only. Travel and accommodation expenses will not be covered.