Simular Secures $21.5M Series A to Scale AI Agents for Mac and Windows

Funding Boost and Investor Backing

Simular announced a $21.5 million Series A financing round led by Felicis. The round also included NVentures, South Park Commons, and other existing investors. This infusion follows a prior $5 million seed round, bringing total capital raised to roughly $27 million. Investors such as Basis Set Ventures, Flying Fish Partners, Samsung NEXT, Xoogler Ventures, and angel investor Lenny Rachitsky are also listed as supporters.

Product Vision and Recent Release

The startup focuses on building AI agents capable of operating a computer’s graphical user interface, rather than merely interacting with a browser. The agents can move the mouse, click, and perform actions like copying and pasting data into spreadsheets. Simular launched version 1.0 of its Mac OS agent, making the technology publicly available on that platform.

Windows Expansion and Microsoft Partnership

Simular is working with Microsoft to develop a Windows counterpart to its Mac agent. The company has been accepted into Microsoft’s Windows 365 for Agents program, joining other firms such as Manus AI, Fellou, Genspark, and TinyFish. While a specific timeline for the Windows release was not disclosed, Simular indicated that the Windows version aims to match or exceed the popularity of the Mac offering.

Founders’ Expertise

Co‑founder and CEO Ang Li is a continuous‑learning scientist who previously worked at Google’s DeepMind, where he met co‑founder Jiachen Yang, a reinforcement‑learning specialist. Their backgrounds include research that contributed to Google products, including Waymo.

Technology Approach: Neuro‑Symbolic Agents

Simular describes its core technology as “neuro‑symbolic computer‑use agents,” a hybrid that is not solely dependent on large language models (LLMs). To address the problem of LLM hallucinations—incorrect or fabricated outputs—the company lets the LLM generate deterministic code after a successful task trajectory is identified. This deterministic code can be inspected, audited, and reused by end users, providing a repeatable and reliable workflow.

Early Use Cases and Open‑Source Community

Beta customers include a car dealership automating vehicle identification number (VIN) searches and homeowners’ associations extracting contract information from PDFs. Simular’s open‑source project, currently limited to Mac OS, has been applied to a range of automations, from content creation to sales and marketing tasks.

Outlook

With substantial venture backing, a clear product roadmap for both Mac and Windows, and a technology stack designed to mitigate hallucination risks, Simular positions itself as a notable player in the emerging field of agentic AI. The company’s next steps will likely focus on scaling its Windows agent, expanding its enterprise customer base, and further refining its neuro‑symbolic approach.