We've backed Ambiguous, AI agent teammates for business

Are they AI agents helping my business, humans? We can't tell, it's Ambiguous!

Today’s work tools assume either humans do the work, or AI replaces it. Ambiguous takes a different approach: AI coworkers that operate alongside humans in the same workspace, with real identities (each with Google accounts), shared context, and a full audit trail. Humans and AI collaborate on the same threads across Google Workspace, Slack, and more, making work feel less like automation and more like teamwork.

What stood out to us is Ambiguous’s focus on the agentic operating system for work. Instead of point tools or chatbots, Ambiguous builds AI coworkers that understand intent, respect constraints, learn from experience, and improve over time. Each coworker develops memory, patterns, and mental models, allowing them to handle real operational work like scheduling, coordination, and follow-through with increasing reliability.

Ambiguous is also built to blitzscale (see our scorecard). The product is designed around product-led growth and virality: AI coworkers naturally interact with other humans, spreading through shared threads, emails, and workflows. As teams add more coworkers and expand use cases, the platform compounds in value, moving from individual productivity to team operations and eventually enterprise-wide adoption.

The founding team Ryan Waliany, Philip Lee, and Dan Hsiao all worked together at a prior company Doorstead, brings a rare combination of deep technical skill and operating experience. They are repeat founders and operators who have built and scaled products together before, with backgrounds spanning AI research, product leadership, and high-growth startups. This shared history shows up in their tight execution, clear product instincts, and strong alignment on long-term vision.

We met the Ambiguous team through the a16z Speedrun program and were immediately impressed by their clarity of thought and ambition. We believe Ambiguous is defining a new category: not just AI assistants, but true AI coworkers that work like real teammates. We’re excited to partner with the team as they build the operating system for agentic work.