Simulatestartup that creates AI agents for Mac OS and Windows has raised $21.5 million in Series A funding led by Felicis, with participation from NVentures (Nvidia’s enterprise arm), existing seed investor South Park Commons and others.
Simular is an interesting agent startup because, unlike others, it doesn’t try to control the browser, but the computer itself. (Agentic AI refers to systems that may autonomously perform complex tasks with minimal human intervention.) “We can literally move the mouse around the screen and click. So it’s able to do more, repeating any human actions in the digital world,” co-founder, CEO Ang Li told TechCrunch, giving an example of copying and pasting data into a spreadsheet.
On Monday, it announced the release of version 1.0 for Mac OS. But it is also working with Microsoft to develop an agent for Windows. The startup is one of 5 agent firms accepted into the Windows 365 for Agents program Microsoft announced this in mid-November. (The others are Manus AI, Fellou, Genspark, and TinyFish.) As for the timeline for the Windows version, Li was vague, apart from to say it might be as popular or more popular than the Mac version.
Another reason to watch Simular is the founders’ good faith: Li is a continuous learning scientist who previously worked at Google’s DeepMind, where he met his co-founder, reinforcement learning specialist Jiachen Yang. Although their team published a significant proportion of papers, Li said their work was not strictly academic in nature. It was intended to improve Google products, including Waymo.
This knowledge of AI products is helpful because many technical issues need to be solved before Silicon Valley’s agentic future can turn into a reality. One of the biggest ones is that LLMs hallucinate a certain percentage of the time.
Agent tasks may require hundreds or tens of millions of distinct steps. Not only can hallucinations at any single step invalidate the agent’s entire work, but hallucinations turn into statistically more likely as the variety of steps increases.
One way to solve this problem is to make the “non-deterministic” LLM “deterministic”, which implies that as an alternative of allowing the LLM to be infinitely creative, its reactions and actions are planned the same way every time. However, this risks limiting the entire creative problem-solving aspect of the agent.
Techcrunch event
San Francisco
|
October 13-15, 2026
Simular marries these two. Its agent will freely iterate through the task, and the user in the middle will revise until the agent succeeds. The human then blocks the workflow of that task, making it deterministic and repetitive.
“Our solution is that agents can continue to explore a successful trajectory. Once a successful trajectory is found, it becomes a deterministic code,” explains Li.
The startup can do this because its work – which Li admits is still early – doesn’t just rely on the LLM wrapper that uploads and downloads data to the model.
“We have a new technology that is not used by any other agent company. We call it ‘neurosymbolic computer-based agents’. It is not fully LLM-based,” he said. “Our approach to solving hallucinations is to allow LLM to write code that becomes deterministic. So if you have a workflow that works, the next time we run the same workflow, we will also be successful.”
Another profit is that this deterministic code that performs a repetitive task is in the hands of the end user, not the LLM. “Once they know the code, they can trust it because they can check it, they can audit it and see what’s going on,” Li says.
Time will tell if this method is the magic that may put agents in the hands of every worker. Li says his early beta customers include automotive dealerships automating VIN lookups and HOA agencies pulling contract information from PDF files. And corporate open source project (currently only available for Mac OS) has led to automation from content creation to sales and marketing.
Simular previously raised $5 million, bringing its total to roughly $27 million. Other investors in the company include Basis Set Ventures, Flying Fish Partners, Samsung NEXT, Xoogler Ventures and podcaster and angel investor Lenny Rachitsky, the company said.