NUA COMPUTING

The industry is currently stuck in binary debates: Local vs. Cloud, Open vs. Closed, or even AI vs. not-AI. The reality is perhaps somewhere in between.

Just because a model can be run locally today doesn't make it useful for general folks; raw inference isn't a useful product for many. We need to bridge the gap between "here's an open model you can run on expensive hardware" and "here's how to do more with your existing devices."

Even as model capabilities keep getting better for the edge, we still need cloud infrastructure and APIs for certain tasks. We need edge-first hybrid systems that deliver privacy where it matters, and power where it is needed. However, local-first hybrid AI today is missing some core engineering foundations that make it more accessible, trustworthy, and usable:

• Hybrid Edge-cloud Foundations (Runtimes, Tooling & Orchestration): Building optimized runtimes, necessary tooling for development and monitoring, reliable tool-calling, orchestration etc. on the edge.
• On-Device Personalization: Engineering ways for AI to learn user preferences and context in resource-constrained environments, ensuring personalization that doesn't require a trade-off with privacy and data sovereignty.
• Interactions: Moving beyond chatbots to explore new HCI primitives — like drawing or other embedded interactions — that feel context-aware, native, and seamless. Some of these interactions actually feel long overdue, even without AI behind them.

Nua is an AI engineering and product lab currently focused on building these missing foundations. These areas of focus are a result of real-world learnings from building and shipping AI products, whether it's around 'Computer Use' or other multimodal interactions. Some of the early product explorations didn't go as expected but those failures have highlighted what's missing.

Ultimately, products are what end-users care about, but products can only be as good as the infrastructure they sit on. The goal is to build these missing pieces of infrastructure so that when products ship, they are genuinely useful and usable by many more.