IrisGo Raises $2.8M to Build an AI Desktop Agent That Learns and Automates User Workflows

IrisGo, a startup building an AI-powered desktop companion for PCs, closed a $2.8 million seed round in 2026 led by Andrew Ng’s AI Fund, with additional backing from Nvidia and Google. The company has since launched beta versions of its macOS and Windows apps and struck a preinstallation deal with Acer.

The application, called Iris, watches a user’s desktop activity and learns to replicate tasks with little to no repeated instruction. Show it a workflow once — such as placing a coffee order or processing an invoice — and it can carry out that same sequence independently going forward. The app also ships with a built-in library of ready-to-use automated workflows covering email drafting, invoice processing, report building, and document summarization, among others. A coding assistant aimed at developers is included as well.

Iris was co-founded by Jeffrey Lai, a former Apple engineer who worked on the Chinese language version of Siri. Lai connected with Ng through a shared alumni network at Carnegie Mellon University, where both studied. After demoing the product to Ng directly, Lai secured AI Fund’s lead investment.

“Our target audience is knowledge workers — white collar companies. There’s a lot of repetitive tasks that those workers do every day,” Lai told TechCrunch, adding that the goal is to shift humans toward high-level conceptual work while agentic systems handle clerical tasks in the background.

On privacy, Iris is designed to process much of its data on-device. Lai describes the architecture as a hybrid system, where more complex tasks are routed to the cloud only with explicit user authorization and under end-to-end encryption.

Beyond the Acer deal, IrisGo says it is pursuing preinstallation agreements with additional laptop manufacturers, suggesting the company may look to grow its user base through hardware partnerships rather than relying solely on direct downloads.