IrisGo Debuts Proactive AI Agent to Automate Knowledge Worker
Led by former Apple engineer Jeffrey Lai, the startup seeks to move AI beyond chat-based prompts toward autonomous task execution in professional
AI reporter
Reports on model launches, frontier labs, developer platforms, and AI policy with an emphasis on claims verification and rollout context.
Editorial responsibility: Lead reviewer for AI coverage, launch claims, and policy context
Primary source: TechCrunch AI. Full source links and update notes are below.
Fast summary
Start here
- IrisGo closed a $2.8 million seed round led by Andrew Ng’s AI Fund, with additional backing from Nvidia and Google.
- The desktop application learns from user behavior to automate complex tasks like invoice processing, report building, and coffee orders.
- The startup has secured a deal with Acer to preinstall the application on new laptops as it expands its distribution.
What happened
IrisGo, a startup backed by Andrew Ng's AI Fund and founded by former Apple engineer Jeffrey Lai, has launched a desktop AI assistant designed to move beyond chat and into proactive workflow execution. The company says its software can observe how users perform repetitive tasks, remember the steps, and later carry them out with minimal prompting. That makes IrisGo part of a growing class of AI products trying to shift the industry from question-answer interfaces toward software agents that actually do the work.
What's new in this update
The most notable new element is distribution. IrisGo has already secured a deal with Acer to preinstall the product on new laptops, which gives the startup a potentially valuable route into everyday desktop environments without forcing users to discover it one app download at a time. For an agentic AI tool, preinstallation matters because habit formation is a major barrier. Products that are already present on the device have a much better chance of being tested repeatedly until users trust them with real tasks.
The funding also matters. A $2.8 million seed round led by Andrew Ng's AI Fund, with additional support from Nvidia and Google, gives IrisGo credibility in a market crowded with ambitious automation claims. Investors are not simply betting on another chatbot layer. They are betting on a product that might become a daily operating surface for knowledge workers.
Key details
IrisGo uses what the company describes as a demonstration model. A user shows the system how to perform a task once, and the assistant records the workflow so that it can repeat the process later. The target tasks include invoice handling, report generation, routine ordering, and other operational chores that consume time but follow repeatable patterns.
The assistant also includes a library of skills for more common business actions such as email drafting, summarization, and coding support. That combination is important: one layer handles personalized automation based on user behavior, while another provides reusable capabilities out of the box.
Privacy is a core part of the pitch. IrisGo says it uses on-device processing whenever possible and only sends data to encrypted cloud environments when the user explicitly allows it for more complex work. For enterprise users or privacy-sensitive professionals, that design may be as important as the automation itself.
Background and context
The broader AI market has spent years training users to type prompts into chat windows. But many companies now believe the next step is not better conversation, but better execution. Desktop assistants that can see context, navigate interfaces, and repeat workflows are one of the most obvious expressions of that trend.
Jeffrey Lai's Apple background is relevant here, especially because he previously worked on Chinese-language Siri. IrisGo's name itself is a reversal of "Siri," which is not subtle. The implication is that voice assistants solved the problem of access, but not the problem of agency. IrisGo wants to be useful not because it speaks, but because it acts.
Competition in this category will be intense. OpenAI, Anthropic, Microsoft, and others all have ambitions around agentic software. IrisGo's bet is that local desktop integration, teach-by-demonstration workflows, and preinstalled distribution can help it stand out.
What to watch next
The real test is whether users trust the assistant enough to delegate meaningful work. Plenty of AI products can demonstrate automation in a controlled environment. Far fewer can handle messy real-world desktop workflows without breaking, misclicking, or requiring constant correction.
The Acer partnership will also be important to watch. If IrisGo can turn preinstallation into sustained active use, it gains an advantage many startups never get. If users ignore it or disable it, distribution alone will not matter.
More broadly, IrisGo is a useful signal about where the AI market is heading. The question is no longer only which assistant can answer best. Increasingly, it is which assistant can watch, learn, and execute routine work with enough reliability that users stop thinking of it as a feature and start treating it as a co-worker.
Why it matters
Proactive AI agents represent a shift from reactive chat interfaces to autonomous systems that anticipate and fulfill professional workflows, potentially reducing clerical overhead for knowledge workers.
Read next
Follow this story through the topic hub, more ai coverage, and the latest updates.
Weekly briefing
Get the week's key developments in one concise email.
Get a fast catch-up on the biggest stories, the context behind them, and the links worth your time.
Cadence
Weekly, for a quick catch-up
Coverage
AI, business, world, security, sports
Format
Clear takeaways and useful context
Request the briefing
Leave your email to open a prepared request and get on the list for the weekly briefing.
About the byline
AI reporter
Alex Rivera reports on artificial intelligence with an emphasis on model launches, frontier lab strategy, developer tooling, and the policy decisions shaping commercial deployment.
Sources and methodology