Proactive systems that anticipate and handle user needs before users recognise them are drawing attention in the AI industry, TechCrunch reported on May 20 local time.

IrisGo is regarded as one of the startups pioneering the field. IrisGo raised $2.8 million in a seed round led by Andrew Ng’s AI Fund early this year.

It is developing a desktop companion app for PCs. It aims to learn users’ daily workflows and run tasks automatically without human instructions.

IrisGo was co-founded by Jeffrey Lai, who helped develop the Chinese version of Siri at Apple. Iris is Siri spelled backwards.

The company says the Iris app remembers a process after being shown a task once and then repeats it automatically. It does not require repeated instructions.

Iris offers a library of “skills” with built-in, ready-to-use automation workflows such as drafting emails, processing invoices, writing reports and summarising documents. It learns users’ desktop behaviour and adds tasks it can automate to a list. It also includes a coding assistant similar to OpenAI Codex or Anthropic Claude Code.

Lai said the target is white-collar knowledge workers. He said frontier models have become powerful but AI-assisted office work is still manual and repetitive. He said the goal is an autonomous workflow that lets people focus on higher-level tasks while an agent handles office work.

Iris processes much of its data on-device, giving it an advantage on privacy compared with cloud-dependent apps. It uses a hybrid setup in which complex tasks go through the cloud.

Nvidia and Google also invested in IrisGo. IrisGo recently launched beta versions of its macOS and Windows apps and is pursuing preinstallation deals with laptop makers for new devices. It has signed a deal with Acer and is in talks with other companies, TechCrunch reported.