Skip to content
NewsProduct launch

OpenAI launches Presence to deploy enterprise AI agents through Forward Deployed Engineers

· by Pondero Newsdesk

The short version

OpenAI launched Presence on July 22, 2026, a managed enterprise service that pairs clients with Forward Deployed Engineers to build narrow AI agents for customer support, IT, and sales workflows.

OpenAI launches Presence to deploy enterprise AI agents through Forward Deployed Engineers

OpenAI's newest product is not software you can sign up for. Presence, announced July 22, 2026, is a managed enterprise deployment service that pairs clients with OpenAI's own engineers to build and run narrow AI agents in production.

What

Presence targets companies that want AI agents handling customer support, sales, and internal IT workflows. Unlike OpenAI's API products, it is not self-serve. Companies must engage through their OpenAI account team; Forward Deployed Engineers and select systems integrators then manage the configuration. Each deployment begins as a single-job agent (billing resolution, insurance claims processing, or IT service requests) and expands only after clearing a structured review cycle.

The platform bundles five components: standard operating procedures that define agent behavior, guardrails that block out-of-bounds interactions, an approved-actions registry, a simulation tool for testing edge cases before launch, and an evaluation suite that grades outcomes against policy compliance and tool usage. Post-launch, a Codex-powered loop reviews production sessions, flags gaps, and proposes behavioral updates that human staff must approve before the changes go live. The underlying inference runs on GPT-5.6 models, per SiliconANGLE.

OpenAI reported that Presence already runs its own English-language developer support line and resolves 75% of inbound calls without human intervention, per Help Net Security. The company also said the agent matched help-desk representative response quality within a few weeks of launch. Both figures are OpenAI self-reports and have not been independently audited.

Presence launched in limited general availability. Pricing was not disclosed.

Why it matters

Presence is OpenAI's first managed enterprise deployment service. The Forward Deployed Engineer model follows professional-services logic: OpenAI staff configure, test, and improve the agents on the client's behalf rather than selling API access and leaving integration to the buyer. That is a different revenue motion from the platform products that built the company.

For operators weighing deployment options, Presence signals that OpenAI is pursuing large-enterprise accounts where self-serve tooling has not closed contracts. The tradeoff is explicit: a white-glove deployment is slower to stand up and less flexible than building directly on the API, but the built-in evaluation loop and guardrails reduce the compliance overhead that stalls many enterprise AI rollouts.

The 75% call-deflection figure, if it holds in third-party deployments, would shift ROI calculations for high-volume support operations. Operators in insurance, finance, and SaaS support should treat it as a hypothesis to test against their own cost-per-ticket baselines, not a purchasing commitment.

What to watch next

The first public case studies from non-OpenAI customers will determine whether the 75% deflection rate travels outside OpenAI's own help desk. A second signal to track: whether OpenAI opens a lighter self-serve path for mid-market accounts that cannot commit to a Forward Deployed Engineer engagement. The managed enterprise AI agent category is attracting competing programs, and second-half contract wins will show whether customers prefer white-glove deployment or platform-led approaches.

Sources