Simular’s AI assistant is designed to operate your Mac or Windows computer on your behalf

Simular Secures $21.5 Million in Series A Funding for AI Agent Development

Simular, a company focused on creating AI agents for both Mac OS and Windows platforms, has successfully closed a $21.5 million Series A funding round. The investment was led by Felicis, with participation from NVentures (Nvidia’s venture capital division), South Park Commons, and several other backers.

Innovative Approach to Agentic AI

What sets Simular apart from other agentic startups is its ambition to control the entire computer environment, not just the web browser. Agentic AI refers to artificial intelligence systems capable of autonomously handling complex tasks with little human oversight. According to CEO and co-founder Ang Li, Simular’s technology can physically move the mouse and perform clicks, enabling it to replicate a wide range of digital tasks typically carried out by humans, such as transferring data into spreadsheets.

Product Launches and Collaborations

The company recently introduced version 1.0 of its Mac OS agent and is collaborating with Microsoft to develop a Windows version. Simular is among five startups chosen for Microsoft’s Windows 365 for Agents initiative, alongside Manus AI, Fellou, Genspark, and TinyFish. While a specific release date for the Windows agent has not been disclosed, Li hinted that it is expected to match or surpass the popularity of the Mac version.

Expertise Behind Simular

Simular’s founders bring impressive credentials to the table. Ang Li, a scientist with a background in continuous learning, previously worked at Google’s DeepMind, where he met co-founder Jiachen Yang, an expert in reinforcement learning. Their work at DeepMind was geared toward practical improvements for Google products, including Waymo, rather than purely academic research.

Addressing Technical Challenges in Agentic AI

The team’s experience is crucial as they tackle significant technical hurdles facing agentic AI. One major issue is the tendency of large language models (LLMs) to generate inaccurate or “hallucinated” outputs. Since agentic tasks can involve thousands or even millions of steps, a single error can compromise the entire process, and the likelihood of mistakes increases with task complexity.

Balancing Creativity and Reliability

One potential solution is to make LLMs deterministic—ensuring they produce the same output for the same input every time. However, this approach can stifle the creative problem-solving abilities of AI agents. Simular’s strategy combines both flexibility and reliability: the agent is allowed to experiment and iterate on tasks, with human users providing guidance as needed. Once a successful workflow is established, it is locked in, making it repeatable and dependable.

“Our approach lets agents explore different paths to success. When a successful method is found, it becomes deterministic code,” Li explained.

Unique Technology and User Empowerment

Simular’s technology goes beyond simply wrapping LLMs. The company has developed what it calls “neuro symbolic computer use agents,” a novel approach not found in other agent startups. This system allows the LLM to generate code that, once proven effective, can be reused reliably. Importantly, this code is accessible to users, who can review, audit, and trust the automated processes.

Real-World Applications and Open Source Initiatives

It remains to be seen whether Simular’s method will become the standard for workplace automation. Early adopters include car dealerships automating vehicle identification number (VIN) lookups and homeowners’ associations extracting contract details from PDF documents. Simular’s open source project, currently available for Mac OS, has enabled automations in areas such as content creation, sales, and marketing.

Funding and Investors

Prior to this round, Simular had raised $5 million in seed funding, bringing its total capital to approximately $27 million. Additional investors include Basis Set Ventures, Flying Fish Partners, Samsung NEXT, Xoogler Ventures, and angel investor Lenny Rachitsky.