Phaze raises $8.75M to give AI agents remote access to workstations
The startup aims to solve the problem of idle, expensive hardware by creating a shared pool of compute for both humans and autonomous AI agents.
The short version
- Phaze raised $8.75 million in seed funding from Andreessen Horowitz, Index, NextView, Makers, and Lerer Hippeau.
- The platform allows AI agents to navigate operating systems and launch apps like humans do, rather than using standard APIs.
- The technology turns individual workstations into a shared pool of elastic compute to prevent expensive hardware from sitting idle.
- The system supports human-to-agent collaboration, such as an AI bot handing off a complex IT ticket to a human engineer.
- Workstations can be located in data centers, on desks, or in storage while remaining accessible via the Phaze streaming tech.
Why it matters
This shifts how companies manage high-end hardware by making it a shared resource instead of a one-to-one assignment. It also provides a necessary infrastructure for AI agents to perform complex tasks that require a full desktop environment.
Solving the idle hardware problem
Many companies invest in high-end workstations that can cost more than $10,000. These machines are necessary for rendering complex 3D graphics or compiling massive codebases, but they are often underutilized. Current IT policies frequently assign one machine to one employee, meaning the hardware sits idle when that person is in a meeting or away from their desk.
Phaze aims to fix this inefficiency by transforming these individual workstations into a shared pool of elastic compute. Instead of being tied to a single person, the hardware can be accessed by anyone in the organization who needs it. This allows companies to get the maximum possible value out of their existing expensive infrastructure.
The startup was founded by the team behind Parsec Technologies Inc., a firm with a history in remote desktop software. By leveraging their experience, Phaze is creating a system where the physical location of the computer does not limit its utility. Workstations can be kept in a data center or even locked in a cupboard while remaining fully accessible to the team.
A desktop for autonomous agents
As companies begin to deploy autonomous AI agents, these programs need a way to perform work without interrupting human users. Many current computer-use agents tend to hijack a worker's screen to automate tasks, which can reduce overall productivity. Phaze provides a dedicated way for these agents to have their own computer access without interfering with human workflows.
The platform uses proprietary low-latency streaming technology to provide high visual fidelity and responsiveness. This allows AI agents to log into and operate expensive workstations in the same way a human employee would. Rather than relying on brittle application programming interfaces (APIs), the agents can navigate the operating system and launch various applications directly.
To make this possible, Phaze developed a specific Model Context Protocol server. This server acts as the bridge that allows the AI to interact with the hardware effectively. By giving agents a
Leveraging a decade of experience
The team behind Phaze is not new to the remote desktop space. The founders have spent more than a decade building technology in this field, which provides them with deep expertise in remote access. This long history allows them to bring established technical knowledge to the current era of machine management and AI integration.
This experience is central to the platform's ability to deliver low-latency streaming. By combining years of remote desktop development with new requirements for AI agents, the company aims to provide a reliable infrastructure for complex enterprise workflows.
Scaling the go-to-market strategy
With the recent seed funding, Phaze is now focused on expanding its reach to the enterprise market. The company plans to use the capital to accelerate its go-to-market plans and begin onboarding new customers. This phase involves moving from a new startup offering to a widely adopted business solution.
To support this growth, the company will also look to expand its engineering team. These new hires will help develop additional features to make the platform more robust for various organizational needs. The goal is to convince businesses of the benefits of dynamic resource allocation over static hardware assignments.
Expanding the infrastructure reach
The platform is designed to be flexible regarding where the actual hardware is physically located. Because the streaming technology works regardless of the machine's spot, organizations can house workstations in data centers or even keep them in storage. This flexibility ensures that the compute power is available whenever a human or an agent needs to perform a task.
By making these machines accessible from anywhere, Phaze helps administrators manage resources more efficiently. This shift allows for a more fluid way of handling fluctuating workloads, ensuring that high-end hardware is utilized to its full potential across the entire organization.
What to watch
- How many enterprises will adopt shared workstation pools over traditional 1:1 hardware assignments.
- The growth of Model Context Protocol servers as a standard for AI-to-computer interaction.
Try it
Look into your company's hardware policy to see if high-end workstations are assigned to individuals or shared as a pool.
Sources and links
- OfficialAI agent workers phaze.app
- Officialboth humans phaze.app
- ReportingSiliconANGLE: the article we sourced
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.
Listen to the AI Wire headlines
Listen to the AI WireListen to our daily AI shows every morning: on your commute, in the shower, wherever you are
Listen to the daily showsFirst reported by SiliconANGLE · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
