Simular, a pioneering startup focused on developing artificial intelligence agents for Mac OS and Windows, has announced the successful completion of a $21.5 million Series A funding round led by Felicis. This extraordinary investment follows the company’s $5 million in previous seed funding. With their most recent $22 million Series C round, their total capital raised climbs to around $27 million. The funding will advance the development and commercialization of Simular’s cutting-edge AI solutions.
Simular was founded by Ang Li, a continuous learning scientist who previously worked at Google’s DeepMind. Together with Jiachen Yang, a reinforcement learning specialist, the company is one of five companies chosen for the Windows 365 for Agents program. This acceptance is a major reflection of the startup’s tremendous impact on the technology landscape. It is committed to improving the user experience with smart automation.
Simular’s AI agents are more than just glorified wrappers around large language models (LLMs). They function as all-in-one solutions that are able to run standardized processes right on the end user’s device. In contrast to most classical AI applications that only act through a web browser, Simular’s agents have the power to take charge of the whole PC platform.
Simular’s technology addresses the key challenges in deploying AI at scale. In particular, it seeks to tackle hallucinations — when AI produces false or random responses. Our goal in addressing hallucinations is to allow the LLM to author code which then makes things deterministic, said Li. The beauty of a good workflow is that when it’s done, it will work again the next time you run it. Have faith in all that’s been accomplished!
This powerful methodology is what enables users to have confidence in the outputs generated by Simular’s intelligent agents. Once they get the code, they should be able to trust it. They can poke at it, they can audit it, they can see precisely what’s happening,” Li continued.
With its full version 1.0 recently released for Mac OS, Simular has already started to draw interest from initial beta customers. Notable uses include a car dealership that automates Vehicle Identification Number (VIN) searches and homeowners’ associations (HOAs) that extract contract information from PDFs. Taken together, these examples represent the breadth and depth of how Simular’s technology is being used in practical real-world situations.
Simular’s capabilities go far beyond simple automation. Its agents are capable of reproducing human behavior in the digital world. Now we could actually physically move the mouse on the screen and do the click. So it’s much better at doing, duplicating redoing everything humans do, say in the real world,” Li said. Such high degrees of functionality have Simular becoming one of the most powerful players in the AI realm.
From today’s cutting-edge innovations like ChatGPT to Ford’s CoPilot, generative and non-generative AI technology is transforming the way users engage with their devices. By enabling direct control over systems rather than just data retrieval, Simular is poised to redefine user expectations and experiences across various industries.
