The OpenAI team has introduced Presence, a platform designed for deploying AI agents in corporate workflows. This tool enables companies to create autonomous assistants for customer interactions and internal tasks.

Presence combines language models with control systems and security protocols. Unlike standard chatbots, these new agents can access internal company databases, perform authorized actions within the system, and escalate conversations to a human operator when necessary.

Developers highlighted several key features of the platform:

  • Access Control: The agent receives only the data and permissions required for specific tasks (e.g., processing refunds or handling insurance claims);
  • Automatic Improvement: The Codex system analyzes gaps in conversations and suggests updates for the algorithms;
  • Multimodality: Presence supports real-time interactions in both chat and voice channels.

OpenAI has already implemented Presence in its English-language phone support service. According to the company, the AI independently resolves 75% of incoming requests. The use of Codex has reduced the frequency of calls transferred to human agents by 15% within 10 days.

Among the first testers of the technology are BBVA bank, telecom giant SoftBank, and insurance group IAG. These companies are utilizing Presence to automate banking operations and assist customers during peak times.

At launch, the product is available to a limited number of corporate clients. Implementation is being handled by OpenAI engineers and certified partners.

In July, Bloomberg reported on the company's plans to enter the consumer electronics market with a "home AI companion" in the form of a smart speaker.