Twin1 AI emerges from stealth with $20M in funding to give every professional an AI-powered digital twin

Twin1 AI is betting that the next big enterprise AI interface won’t be another chatbot. It will be a digital version of you.

The San Mateo, California-based startup emerged from stealth Thursday with $20 million in seed funding to build AI-powered digital twins that capture a professional’s knowledge, judgment, work context and communication style. The round was co-led by Bessemer Venture Partners, Tribeca Venture Partners and Aramco Ventures.

Founded in 2025 by Dr. Lewis Z. Liu, Tom Cahn, Huiting Liu and Dr. Jonathan Budd, Twin1 AI has teams in San Mateo and London. The startup plans to use the fresh capital to grow both locations, invest in go-to-market operations and continue developing its technology.

“The funding will be used to expand the company’s teams in San Mateo, CA and London, UK, invest in go-to-market efforts, and further develop its core technology,” Twin1 said in a statement.

Twin1’s pitch starts with a problem that has become increasingly visible as companies deploy AI agents: much of an organization’s valuable knowledge isn’t sitting neatly inside a database. It lives across people’s emails, meetings, documents, conversations, and years of accumulated experience.

Twin1 wants to make that human context accessible to AI without stripping employees of control.

A digital twin that works where you work

Each user gets an AI digital twin that learns from workplace context, including email, meetings, documents and enterprise systems. The twins can operate across Slack, Microsoft Teams, Outlook, Gmail, Google Drive and SharePoint, answering questions and acting on behalf of the people they represent.

The company says users retain control over what information their twin can access and how that information can be shared. Twin1 combines enterprise policies, existing permissions, AI-based controls and human approval to govern exchanges between employees, their twins and other AI systems.

That distinction could matter as companies move from AI tools that generate answers to agents that can take action. Giving an agent access to corporate data is relatively straightforward. Giving it the unwritten context behind who knows what, who can see what, and how decisions get made is a much harder problem.

Twin1 calls its approach the “Twin Network.” Individual twins can locate expertise across an organization, request permission-aware information, and coordinate work without taking control away from the humans they represent. An enterprise MCP server lets outside AI agents and workplace applications access governed context from individual twins or the broader network.

“In every knowledge organization, the human is the atomic unit of knowledge. The opportunity with AI is not to flatten people’s expertise into generic output, but to connect and amplify their unique knowledge, judgment and context,” said Dr. Lewis Liu, co-founder and CEO of Twin1 AI. “Our belief is that AI should work for people, not be done to them. Twin1 gives every professional a digital twin that preserves and amplifies their knowledge, judgment, and voice, while keeping privacy, permissions, and human agency at the core. We are building a future where expertise compounds, rather than collapses into the average.”

Twin1 isn’t starting from zero. The company says it has deployed its technology with partners in legal services, finance, and energy for more than a year. Customers include Linklaters, Orrick, Dechert, Customers Bank and Aegis Energy. Twin1 says customers report automating 30% to 50% of the communications work knowledge workers do.

“Enterprise knowledge has become increasingly fragmented, creating real friction in how organizations access and use expertise,” said Elliott Robinson, partner at Bessemer Venture Partners. Twin1 AI is building an individual-first context layer that allows users to extend their knowledge across teams and systems. We believe this represents an important shift in how AI can be applied to knowledge work while preserving control and ownership.”

The round drew a long list of other investors, including EJF Ventures, Tin Alley Ventures, AGI House Ventures, Neo, F-Prime, Btech Consortium, Antiportfolio Ventures, Lakestar, Notion Capital and Insiders, plus a strategic investment from Orrick and several angel investors.

Some investors previously backed Eigen Technologies, the earlier company associated with members of Twin1’s founding team.

The bigger question now is whether digital twins can become a new layer of enterprise software. AI assistants know what users tell them. Twin1 is betting that useful enterprise AI needs something deeper: a permission-aware representation of the person behind the prompt.