Mirendil has successfully raised $200 million in a seed funding round led by a16z and Kleiner Perkins, with additional participation from NVIDIA. The company is developing a self-accelerating AI system designed for research and development, which operates autonomously to address complex engineering problems without human input. This innovation reflects a growing trend among AI “neolabs” aiming to democratize access to frontier AI technologies, allowing various enterprises and research groups to manage their own AI endeavors independently rather than relying on a few dominant labs. Mirendil’s founding team comprises 20 experienced researchers and engineers from prestigious institutions such as Anthropic and Google DeepMind, united with the goal of accelerating scientific progress.

a16z: a16z (Andreessen Horowitz) is a Silicon Valley venture capital firm that invests in technology companies across stages, with a strong focus on AI, crypto, and software platforms. In this story, a16z is highlighted as one of the lead investors in Mirendil’s substantial seed round and welcomes the company into its portfolio.
NVIDIA: NVIDIA is a leading technology company best known for its GPUs and AI computing platforms, which power many of the world’s most advanced AI models and data centers. In this news, NVIDIA is described as making a major follow-on investment in Mirendil’s seed round, underscoring the hardware and infrastructure relevance of Mirendil’s autonomous AI R&D systems.
Mirendil: Mirendil is a frontier AI research and product company building autonomous AI systems that can conduct AI research and engineering work, including running their own GPU infrastructure. In this news, Mirendil announces a large seed round to develop a self-accelerating AI R&D agent designed to help organizations run their own advanced AI research loops rather than relying solely on large centralized labs.
Bneyshabur: Behnam Neyshabur is an AI researcher and entrepreneur known for his work on deep learning and generalization, and he is a co-founder of Mirendil. In this announcement, Neyshabur introduces Mirendil publicly, explains its mission of self-accelerating AI R&D, and confirms the funding led by a16z and Kleiner Perkins with participation from NVIDIA.
Harsh Mehta: Harsh Mehta is an AI engineer and co-founder of Mirendil with prior experience at leading frontier AI labs such as Anthropic, Google DeepMind, or xAI as referenced by the team’s background. In the news, he is named as one of the core co-founders building Mirendil’s autonomous AI research system and is recruiting researchers and engineers to join the effort.
Tara Rezaei: Tara Rezaei is a co-founder of Mirendil and a member of its founding research and engineering team with experience at top frontier AI labs. In the news, Rezaei is introduced as one of the leaders responsible for designing Mirendil’s system that autonomously iterates on research and engineering problems using its own compute resources.
Shayan Salehian: Shayan Salehian is a co-founder of Mirendil and part of the founding technical team that previously worked at major frontier AI institutions including Anthropic, Google DeepMind, xAI, and OpenAI. In the announcement, Salehian is cited as one of the key builders of Mirendil’s end-to-end self-accelerating AI R&D platform.

`json
{
“Neolab_trend”: “Mirendil is part of a broader wave of so-called AI ‘neolabs’ that are raising significant early-stage funding to build independent frontier model labs outside the largest incumbents.”,
“AI_RnD_agents”: “Industry discussions over the past month highlight growing interest in autonomous AI agents focused on software engineering and research workflows, with Mirendil positioning its system as a specialized agent for AI R&D that manages its own compute and experimentation loop.”,
“Frontier_AI_competition”: “Recent commentary on emerging labs notes that new players like Mirendil aim to challenge the dominance of a few big AI labs by offering more customizable, self-hosted frontier AI capabilities to enterprises and research groups.”
}
`