OpenAI shifts to high-touch enterprise model with Presence agent platform

OpenAI has introduced Presence, a product designed to help enterprises deploy and manage real-time voice and chat agents. The platform is available immediately through a limited general availability program, but it is not offered on a self-service basis.

The system packages the policies, guardrails, system connections and evaluation tools required to run AI agents reliably. It restricts agents to specific tasks and system access, allowing companies to test behavior against edge cases before launch and monitor performance afterward.

When operational issues arise, OpenAI’s Codex model investigates performance signals and proposes updates, which teams test against the live version before deploying. OpenAI reports that Presence handles 75% of inbound issues on its own English-language phone-support line, and its Codex improvement loop reduced human handoffs by 15 percentage points over a 10-day period.

Several large organizations are testing the foundation, including BBVA for routine Mexican banking voice support and Australian insurer IAG for high-demand periods like natural disasters. “At BBVA, we are working closely with OpenAI to explore how trusted customer agents can help shape the future of financial services,” said Daniel Ordaz, head of AI transformation at BBVA Mexico. SoftBank is also testing Japanese-language conversations, with vice president Tadahisa Murakami noting the platform must "represent SoftBank consistently across customer interactions."

For investors and market professionals, the significance lies in OpenAI’s go-to-market strategy. Presence relies on OpenAI Forward Deployed Engineers and select global systems integrators, a delivery method that closely resembles the model pioneered by Palantir. While Palantir focuses on data ontologies and operational decisions, OpenAI is applying this high-touch, on-site model specifically to AI agent behavior, pushing the company into more labor-intensive, services-heavy revenue streams.

The rollout mirrors a broader industry realization that enterprises need hands-on assistance to manage deployment risks, security and compliance. Anthropic launched a similar consulting organization, Ode, just a week ago. OpenAI also formed its own integration firm, the OpenAI Deployment Company, in May 2026 with backing from Bain & Company.

Critical financial details remain unresolved. OpenAI has not disclosed pricing, geographic limits, contractual terms or the expected cost of the required engineering work. The company also has not said whether Presence can integrate models from external providers, leaving questions about total cost of ownership for prospective enterprise customers.