Hark

Hark is a San Jose-based AI hardware company founded by Brett Adcock, the entrepreneur behind Vettery, Archer Aviation, and Figure AI. The company is building what it calls personal intelligence by combining proprietary multimodal AI models with purpose-built hardware designed to act as a new interface between people and machines.

In May 2026, Hark announced a Series A funding round of more than $700M at a $6B valuation. The round was led by Parkway Venture Capital and included NVIDIA, AMD Ventures, Intel Capital, Qualcomm Ventures, Salesforce Ventures, Brookfield, ARK Invest, Greycroft, Prime Movers Lab, Align Ventures, and Tamarack Global.

Hark operates at the intersection of AI hardware, multimodal AI models, agentic systems, and consumer computing. The company matters because it is testing a large and uncomfortable thesis: the next major AI interface may not be another app. It may require a new device, a new model layer, and a new operating relationship between humans and software.

For investors, operators, and technology leaders, Hark is becoming a high-profile test case for whether vertically integrated AI companies can create the next generation of consumer computing experiences.

About Hark

Most AI startups begin with a chatbot. Hark started with a more expensive question. What happens when artificial intelligence is no longer trapped inside devices that were never designed for it?

That question sits at the center of Hark's strategy. Founded in 2025 by Brett Adcock, Hark is developing a vertically integrated AI platform that combines foundation models, agentic software systems, and custom hardware into a single experience. Unlike companies building AI applications on top of existing operating systems, Hark is designing the stack together.

Hark uses the term personal intelligence to describe AI systems that maintain persistent context, memory, and awareness across a user's daily digital and physical interactions. That ambition places the company somewhere between an AI lab, a hardware startup, and a next-generation computing platform.

The mission is simple enough to fit in a sentence: bring advanced personal intelligence to everyone in the world. The difficulty is in everything after the sentence.

Why Hark Matters Right Now

The AI market has entered an interesting phase. For the last several years, the dominant conversation centered on model performance. Which model was smarter? Which benchmark was higher? Which company had access to the most compute? Now a different question is getting louder. What is the device for AI?

Hark exists because many industry leaders believe today's hardware is fundamentally mismatched with the capabilities of modern artificial intelligence. Smartphones were built for apps. Laptops were built for software. Neither was designed around persistent, multimodal AI agents capable of understanding speech, vision, memory, and context at the same time.

This belief helps explain why investors committed more than $700M to a company that has not yet publicly launched its product. They are not underwriting a feature. They are underwriting a platform thesis.

The presence of NVIDIA, AMD Ventures, Intel Capital, and Qualcomm Ventures in a single round is unusual. It signals broad conviction across the semiconductor ecosystem that AI hardware and personal computing interfaces are becoming strategic terrain.

The Problem Hark Is Solving

Current AI systems remain surprisingly fragmented. They can answer questions. They can generate content. They can write code. What they generally cannot do is maintain a deep, persistent understanding of a person's life across contexts, devices, and interactions.

According to public descriptions of its product vision, Hark aims to address that gap by creating personal intelligence that remembers, adapts, and interacts naturally through speech, vision, and context awareness. The company is building multimodal foundation models internally rather than relying entirely on third-party providers. Just as important, Hark intends to pair those models with dedicated hardware. That combination matters.

Technology history repeatedly shows that software breakthroughs often require new hardware experiences before they reach mass adoption. The mouse helped unlock graphical computing. The smartphone helped unlock mobile internet. Hark believes personal AI may require its own native interface. Whether that belief proves correct remains one of the more interesting open questions in technology.

Leadership and Team

The strongest signal around Hark may be the people choosing to build it. Brett Adcock is not a first-time founder chasing a trend. His career spans recruiting technology, aerospace, robotics, and now AI hardware. Vettery exited to Adecco. Archer Aviation pushed electric flight into the public markets. Figure AI became one of the most visible names in humanoid robotics.

Hark also recruited Abidur Chowdhury as Head of Design. Bloomberg reported that Chowdhury, an Apple veteran associated with iPhone Air design work, joined Hark to lead design. TechCrunch also described his role as Director of Design at Hark.

That hire matters because Hark is not positioning hardware as a sidecar to software. It is treating design as part of the intelligence layer. If the interface is the product, design is not decoration. It is strategy.

Why Hiring Momentum Matters

Many companies hire because they are growing. The most interesting companies hire because they are building something difficult. Hark's careers page shows roles across AI research, AI infrastructure, hardware engineering, embedded systems, operating systems, privacy, security, design, manufacturing, and AI safety. That breadth offers a useful clue about the roadmap.

When a startup recruits operating system architects alongside AI researchers and hardware engineers, it signals a desire to control more of the user experience.

The hiring pattern also reinforces Hark's commitment to vertical integration. Rather than outsourcing critical layers of the stack, the company appears focused on developing internal expertise across the components required to deliver personal intelligence at scale. That is not normal app-company hiring. That is platform-company hiring.

What This Signals for AI Hardware

The investor list behind Hark may be as important as the product itself. It is unusual to see NVIDIA, AMD Ventures, Intel Capital, and Qualcomm Ventures participating in the same financing round. These organizations represent different positions across the AI compute and semiconductor ecosystem. Their collective presence suggests broad interest in whatever computing category emerges after today's AI applications.

Hark also reflects a larger shift happening across the AI landscape. The market is gradually moving from model competition toward experience competition. Raw intelligence still matters. But increasingly, the winners may be companies capable of packaging intelligence into products people actually want to use every day. That challenge is considerably harder than improving benchmark scores.

The Bigger Industry Shift

For now, Hark remains largely a promise. The company has not publicly released its first product. It has not disclosed revenue. It has not announced customers. Public reporting says its first AI models are expected in summer 2026, with hardware planned afterward. That uncertainty is exactly what makes Hark worth watching.

The company sits at the intersection of several powerful trends: multimodal AI, agentic systems, custom hardware, consumer computing, and human-centered interfaces. If Hark succeeds, it could help define what personal intelligence looks like in the coming decade. If it fails, it will still provide valuable lessons about how difficult it is to build entirely new computing platforms. Either outcome will be closely studied across Silicon Valley.