AI Startup Twin1 Emerges From Stealth With $20 Million Funding

AI startup Twin1 AI has emerged from stealth with a $20 million seed round to develop digital “twins” designed to represent individual professionals and carry out work on their behalf.

The San Mateo, California-based company announced the funding on August 20, 2026. The round was co-led by Bessemer Venture Partners, Tribeca Venture Partners and Aramco Ventures. Twin1 said the funding will be used to continue developing its technology, expand its teams in the United States and United Kingdom, and support go-to-market efforts.

Twin1 was founded in 2025 by Dr. Lewis Z. Liu, Tom Cahn, Huiting Liu and Dr. Jonathan Budd. The founding team previously worked together at Eigen Technologies, an enterprise artificial intelligence company that was acquired by SirionLabs in 2024.

Rather than building another general-purpose chatbot, Twin1 is developing an AI system around the knowledge and working context of individual professionals. The company says each user can have a digital Twin that is grounded in approved workplace information such as emails, meetings, documents and other work systems.

According to Twin1, these Twins are designed to answer questions, draft communications, coordinate work and perform tasks within the permissions given to them by their users. The system can operate inside workplace applications including Slack, Microsoft Teams and email.

The company describes the product as a way of extending a professional’s knowledge and judgment rather than simply automating individual tasks. Its central idea is that important organizational knowledge often remains with employees instead of being captured completely in documents or databases.

Bessemer Venture Partners described Twin1 as a “governed context layer” for enterprise artificial intelligence. In its investment announcement, the firm said a Twin is intended to act as a live model of a professional’s expertise and judgment while operating inside the tools they already use. The investor said multiple Twins can be connected to form a network through which employees and AI agents can access information without bypassing the individual who controls that knowledge.

Twin1 calls this system the “Twin Network.” The company says Twins can communicate with one another, allowing expertise from different professionals to be accessed across an organization while remaining subject to permissions and privacy controls.

The platform is being designed for organizations where confidentiality and control of information are important. Twin1 says users and organizations can determine what a Twin can access, who can interact with it and what information it is permitted to disclose. The company also says human approval can be required for certain actions.

Privacy is a central part of the company’s pitch. Twin1 states that customer data will not be used to train, fine-tune or customize an underlying large language model without explicit written consent. It also says the private knowledge of one customer is not used to inform another customer’s deployment.

Twin1 offers several deployment options, including software hosted by the company, single-tenant environments and private-cloud deployments within a customer’s own firewall. The company lists security and compliance credentials including SOC 2 Type II and Google CASA Level 3 on its website.

The startup is initially targeting regulated and knowledge-intensive sectors. Twin1 says it has deployments in legal services, financial services and energy, with named organizations including Linklaters, Orrick, Dechert, Customers Bank and Aegis Energy.

The company’s legal-sector focus is connected to the founders’ previous experience at Eigen Technologies. In an interview with Artificial Lawyer, Liu discussed an earlier problem encountered while working with legal professionals: documents can show the final result of a decision without fully capturing the reasoning, discussions and context behind it.

Liu previously led Eigen, which focused on artificial intelligence for enterprise documents and contracts. Bessemer said Eigen’s technology processed more than $100 trillion in contracts and was used by major financial institutions and law firms before the company was acquired by SirionLabs.

Twin1’s technology also includes an enterprise MCP server intended to allow AI agents and enterprise applications to access information from a user’s Twin under the platform’s permission controls. The company says this allows its digital Twins to work with business systems rather than operating as isolated chat interfaces.

Twin1 says customers have reported that the platform can handle between 30% and 50% of communications work that knowledge workers would otherwise perform themselves. That figure is a company-reported claim and has not been independently established as a benchmark across the startup’s customer base.

The startup has also said it has a pipeline of more than 400 prospects. That figure refers to prospective customers and should not be interpreted as the number of organizations already using or paying for the product.

The funding round included a broad group of institutional and individual investors. In addition to the three lead investors, participants included EJF Ventures, Tin Alley Ventures, AGI House Ventures, Neo, F-Prime Capital, Lakestar, Notion Capital, Btech Consortium and Antiportfolio Ventures. Orrick also participated as a strategic investor.

Named individual investors include Haakon Overli, co-founder of Dawn Capital; Roy Reznick, co-founder of Wiz; Hans Tung, managing partner at Notable Capital; Kevin Buehler, a former McKinsey senior partner; Robert Trezona, a climate-tech investor; and Dan Wong, a former global co-head of Macquarie Capital.

Aramco Ventures’ participation adds to Twin1’s focus on large enterprises and regulated industries. The Saudi oil company’s corporate venture arm joined Bessemer and Tribeca as a co-lead investor in the round, while Twin1 has identified energy as one of the sectors in which it has deployments.

The company’s immediate strategy remains focused on enterprise customers, particularly legal and financial organizations. Liu has said Twin1 intends eventually to make the technology available to smaller firms and individual professionals through a more self-service model.