iGenius, a company specializing in Responsible AI for highly regulated industries such as government, finance, and heavy industry, has announced its rebranding to Domyn. Domyn focuses on providing Large Language Models (LLMs) and Agentic AI solutions designed to bridge the gap between compliance and innovation for organizations operating in the world's most regulated sectors. This transformation, which follows the company's launch of state-of-the-art LLMs and one of the world's largest AI supercomputers earlier this year, aims to empower enterprises with AI they can own, govern, and trust, laying the foundation for what Domyn anticipates as a breakthrough 2025.

Uljan Sharka, Domyn's founder and CEO, described the rebranding as a "natural evolution" of the company's efforts to empower enterprises to "truly own their domain of intelligence." He emphasized that becoming Domyn signifies a deeper commitment to their long-held belief that AI is a powerful force that should be accessible to change-makers and ultimately work to elevate humankind. Sharka stated, "The new name represents a timeless emblem of purpose that proudly reflects our team's mission to uplift and improve the world."

The name "Domyn" is derived from the term "domain," championing the concept of AI sovereignty. It reflects the right for organizations to define their own technological destiny and operate with the autonomy they deserve. According to the company, Domyn isn't about using a new AI; it's about empowering organizations to create their own AI, reinforcing the principle that "Your AI is your IP."

Domyn aims to enable its clients to lead with confidence and accountability by delivering solutions that safeguard intellectual property and ensure end-to-end governance. The company serves some of the world's leading financial service organizations, as well as defense and heavy industry clients, helping them to innovate rapidly, securely, and in full regulatory alignment.

In the government and defense sectors, Domyn supports mission-critical operations, including multilingual threat intelligence triage and secure, air-gapped deployments for decision support. For the finance industry, its AI solutions streamline reporting processes and help combat financial risk through automated compliance analysis. In advanced manufacturing, Domyn powers AI engineering brains through the integration of information technology (IT) and operational technology (OT), and facilitates predictive quality inspection within highly secure environments.

Domyn has been at the forefront of technological advancements, having pioneered large-scale FP8 pretraining. This innovation has successfully reduced inference costs by 50% without sacrificing quality, fueling some of the world's leading LLMs with hundreds of billions of parameters. Furthermore, by advancing Continual Pretraining (CPT), Domyn empowers regulated organizations to build their general intelligence, delivering adaptable, high-performance models grounded in their specific enterprise context.

A core tenet of Domyn's philosophy is the critical importance of owning AI models, especially in a world where generative AI defines strategic advantage. The company likens putting proprietary data into a third-party model to "adding sugar to a cake — once it's baked, it's nearly impossible to separate," stressing that an organization's AI constitutes its intellectual property.

Domyn is dedicated to researching and developing Responsible AI tailored for regulated industries. It offers a composable AI architecture, which includes foundational LLMs built for advanced reasoning and customized to each organization's unique language, logic, and mission. Domyn allows customers to train their own state-of-the-art LLMs, ensuring that businesses truly own their AI and the value it generates. These models power intelligent, customizable agents that can act on private data to perform specialized tasks, all managed securely through a unified platform equipped with built-in tools for access management, auditability, and governance. The company's capabilities are supported by one of the world's largest AI supercomputers, purpose-built for training trillion-parameter models for sovereign, high-stakes applications.