Back to Blog

IrisGo: Andrew Ng-Backed AI Agent That Learns Desktop Workflows

IrisGo: Andrew Ng-Backed AI Agent That Learns Desktop Workflows IrisGo: Andrew Ng-Backed AI Agent That Learns Desktop Workflows IrisGo: Andrew Ng-Backed AI Agent That Learns Desktop Workflows

IrisGo: An AI Desktop Companion That Learns Your Workflows

Overview

IrisGo is a startup backed by Andrew Ng's AI Fund that has raised $2.8 million in seed funding. The company is building a desktop AI agent for PCs that learns user workflows through observation and automates them with minimal human prompting.

Key Features

Learn-Once, Automate Forever

  • One-time demonstration: Show the program a task once, and it remembers the process for future automated use
  • No repeat instructions needed: The system automatically learns from desktop behavior
  • Skills library: Comes with built-in automated workflows for common business tasks:
    • Email drafting
    • Invoice processing
    • Report building
    • Document summarization

Technical Architecture

  • On-device processing: Processes significant data locally for stronger privacy protections
  • Hybrid cloud approach: Complex tasks processed through cloud with end-to-end encryption
  • User authorization required: Cloud processing only occurs when explicitly authorized

Coding Assistant

  • Built-in coding assistant similar to OpenAI's Codex or Anthropic's Claude Code
  • Designed to assist developers during active work

Target Market

Primary audience: Knowledge workers and white-collar companies dealing with repetitive daily tasks

Value proposition: Move from manual, repetitive AI-assisted work toward fully autonomous workflows where humans focus on high-level conceptual work while AI handles clerical tasks

Founding Team & Backing

  • Co-founder: Jeffrey Lai, former Apple engineer who helped build the Chinese language version of Siri
  • Lead investor: Andrew Ng's AI Fund (co-founder of Google Brain)
  • Additional backers: Nvidia and Google
  • Connection: Lai and Ng are both Carnegie Mellon University alumni

Go-to-Market Strategy

Current Status

  • Beta versions launched for macOS and Windows
  • Recently struck preinstall deal with Acer
  • Pursuing additional laptop manufacturer partnerships

Distribution Strategy

  • Direct app downloads
  • Pre-installation on new devices through OEM partnerships

Privacy & Security

  • Strong on-device processing capabilities
  • End-to-end encryption for cloud operations
  • Explicit user authorization required for cloud processing
  • Detailed privacy protections outlined at irisgo.ai/privacy

Industry Context

The company positions itself within the emerging "proactive AI" space—systems that anticipate user needs and fulfill them before explicit requests. This represents a shift from reactive chatbots to autonomous workflow agents.