According to industry experts, the next major advancement in artificial intelligence will be "proactive" systems—agents capable of anticipating and fulfilling a user's needs before the user even recognizes them. One startup aiming to make strides in this space is IrisGo. Earlier this year, the company secured a $2.8 million seed round led by Andrew Ng's AI Fund, and it is developing a desktop companion for PCs that learns a user's daily workflows and automates them with minimal or no human input. Iris was co-founded by Jeffrey Lai, a former Apple engineer who contributed to building the Chinese language version of Siri. (Notably, Iris is Siri spelled backward.)

The fundamental concept is straightforward: demonstrate a task to Iris once, and it retains that process for future automated execution—no need for repeated instructions. During a demonstration, Iris recorded the steps required to order a latte from Philz Coffee, a popular Bay Area chain, including entering credit card details and completing the purchase. Lai then instructed Iris to repeat the order independently, and the agent complied seamlessly. However, buying coffee is not the primary objective. Instead, the system is intended to automate a wide range of business-related activities. Iris includes a built-in "skills" library featuring tasks such as email drafting, invoice processing, report creation, document summarization, and other ready-to-use automated workflows. Additionally, Iris learns from the user's desktop behavior and automatically adds those tasks to its list of potential action items. The application also incorporates a coding assistant, similar in concept to OpenAI's Codex or Anthropic's Claude Code, designed to aid developers in their work.

"Our target audience is knowledge workers—white-collar companies. There’s a lot of repetitive tasks that those workers do every day," Lai explained, noting that despite the advanced capabilities of today's frontier models, AI-assisted office work often remains manual and tedious. The goal, he said, is to shift toward a more fully autonomous workflow, where humans focus on high-level conceptual work while agentic systems handle all clerical tasks in the background. One appealing feature of Iris is its ability to process significant amounts of data on-device, offering stronger privacy protections compared to applications that rely heavily on cloud processing. Lai noted that the system uses a hybrid architecture, meaning larger, more complex tasks are ultimately processed through the cloud. However, the company assures that cloud processing "only occurs when explicitly authorized by the user and uses end-to-end encryption."

Part of IrisGo's scaling strategy involves gaining credibility through associations with prominent figures and organizations. Support from Ng—a co-founder of the influential deep learning research team Google Brain—has been instrumental. Lai arranged a meeting with Ng through a shared connection: both are alumni of Carnegie Mellon University. During that meeting, Lai and his co-founder demonstrated Iris, leading Ng's AI Fund to lead the startup's seed round. Nvidia and Google have also backed the company. IrisGo recently launched beta versions of its macOS and Windows apps, and it is currently pursuing deals with laptop manufacturers to preinstall the application on new devices. The company recently secured such a deal with Acer, and Lai expressed hope that similar agreements with other device makers will follow soon.