Viqus Logo Viqus Logo
Home
Categories
Language Models Generative Imagery Hardware & Chips Business & Funding Ethics & Society Science & Robotics
Resources
AI Glossary Academy CLI Tool Labs
About Contact

Simular Raises $21.5M Series A with Novel Agentic Approach

AI AI Agents Startups Venture Capital Microsoft macOS Hallucinations LLMs Simular
December 02, 2025
Viqus Verdict Logo Viqus Verdict Logo 9
Code-Driven Automation
Media Hype 7/10
Real Impact 9/10

Article Summary

Simular, a new AI agentic startup, has closed a $21.5 million Series A round to accelerate the development of its agents for Mac OS and Windows. Unlike many agentic AI systems, Simular's approach focuses on controlling the entire PC rather than just a browser, allowing for more complex, autonomous task completion. The key differentiator is its ‘neuro symbolic computer use agents,’ which allow the LLM to generate deterministic code based on successful task trajectories, reducing the risk of hallucinations. This deterministic code is then handed over to the end-user for inspection and auditing. The company’s technology, already demonstrated with applications like automating VIN number searches for car dealerships and contract information extraction, is being backed by prominent investors including Felicis, NVentures, and Samsung NEXT. This investment is particularly significant given the challenges currently plaguing the broader agentic AI space, specifically the issue of LLM hallucinations, which often derail complex task automation.

Key Points

  • Simular raised $21.5 million in Series A funding led by Felicis.
  • The startup’s core technology focuses on deterministic agentic AI, mitigating hallucinations by generating executable code from successful LLM task trajectories.
  • Their approach differs from other agentic AI systems by controlling the entire PC, rather than solely the browser.

Why It Matters

The success of Simular represents a crucial step in addressing a major hurdle within the burgeoning agentic AI field: the persistent problem of LLM hallucinations. These hallucinations have rendered many agentic AI solutions unreliable and impractical. Simular’s novel deterministic approach, combined with the backing of established investors and a focus on user-trustable code, suggests a viable path toward more robust and reliable autonomous task completion. This is particularly important for enterprises seeking to integrate AI agents into their workflows, where accuracy and consistency are paramount.

You might also be interested in