Frontier, the new OpenAI platform, connects internal systems and facilitates the management of AI agents, aiming to improve efficiency in companies.

OpenAI has launched Frontier, a platform that functions as a layer of intelligence, connecting disparate systems and databases within organizations. It simplifies the management and building of AI agents, which can autonomously complete tasks on behalf of the user. Frontier reflects an open approach, recognizing that OpenAI cannot develop all types of agents necessary for companies on its own. In recent years, OpenAI has significantly grown in the corporate segment, having over a million business clients, and revenues from this sector are estimated to reach 50% by the end of the year.

The Frontier platform is complementary to existing solutions, such as ChatGPT Enterprise, and allows AI agents to function as virtual colleagues without requiring the restructuring of internal systems. It supports agents developed both by OpenAI and by partner companies like Google and Microsoft. Frontier integrates internal applications, ticketing systems, and data warehouses, enabling agents to run code and interact with the organization's files. It also includes tools for performance assessment, helping to optimize agents in real-time. The initial launch of Frontier is limited to a select set of companies, including Uber and State Farm, with availability expansion scheduled for the coming months.