Presence: A Deployment Definition, Not Another Model
OpenAI Presence is an enterprise AI agent deployment platform that combines real-time voice and chat agents with policies, guardrails, simulations, and ongoing improvement tools so companies can connect agents to internal systems, enforce governance, and run customer-facing and internal workflows in production without exposing raw models directly. OpenAI has announced Presence as a new enterprise product for deploying and managing AI agents across customer-facing and internal business workflows. The debt-fueled maker of frontier models introduced Presence on July 22 as a web service that makes it easier for enterprises to deploy AI agents across common tasks such as customer support, outbound sales, and high-risk internal workflows. This is not a neutral API drop; it is OpenAI’s clearest move yet to sell managed AI services rather than model access.
From Cheap Models To Boots-On-The-Ground Managed AI Services
Presence is a signal that the era of commoditized AI models is giving way to managed AI services. Having popularized AI with low-cost access for the masses, OpenAI is now focusing on enterprise customers who will pay for full deployments, not just tokens. Rather than releasing another self-service API, Presence is delivered through the OpenAI Deployment Company and Forward Deployed Engineers, a tech installation workforce bolstered by the acquisition of consultancy Tomoro. Deployments are led by those engineers and select global systems integrators; the product is explicitly not self-serve. During the limited general availability phase, each deployment is scoped individually based on the customer’s use case and implementation needs. In other words, enterprises are being charged for boots-on-the-ground integration, workflow design, and ongoing management. That makes Presence far closer to a consulting-led platform than a typical developer API—and it repositions OpenAI directly inside enterprise operations.
Guardrails, Governance, And The Real Work Of Enterprise AI Adoption
The OpenAI Presence platform is explicitly designed to solve the unglamorous parts of enterprise AI adoption: governance, security controls, and change management. Presence brings together policies and standard operating procedures, guardrails, approved actions, simulations, evaluation tools, and a Codex-powered improvement process so teams can run agents in production. Together, these components help connect company systems, define how agents should behave, evaluate performance, enforce policies, and manage changes after launch. Presence lets companies set policies and governance controls to determine how agents behave, giving reassurance that voice agent banter and chatbot replies stay polite, professional, and on task. Guardrails can intervene when an interaction moves outside the company’s boundaries. This is where managed AI services earn their keep: they turn abstract model capabilities into accountable workflows with voice agent guardrails, escalation rules, and audit trails that satisfy operational and compliance teams.
Connecting Agents To Enterprise Data—And Owning The Workflow
Presence goes beyond chat interfaces to run the work itself. OpenAI describes it as a deployment platform that supports real-time experiences across voice and chat for customer support, outbound sales, and high-risk internal workflows. A typical customer scenario is a billing issue: Presence can understand the request, verify the customer, look up account information, apply company policy, and take an approved action. According to OpenAI, “Presence already handles its own English-language phone support line and resolves 75% of inbound calls without human intervention.” Each deployment starts with one defined job and gives the agent only the data and system access needed, with company-set policies for approvals and escalation and human review at the edges. Components such as simulations and graders test common requests, edge cases, and higher-risk scenarios before launch, while a Codex-powered loop reviews production interactions and suggests improvements that staff must approve. This is enterprise AI agent deployment as ongoing operations, not a one-off integration project.
Market Shock: Presence And The Economics Of AI Agents
Presence did not land quietly. OpenAI introduced the product on July 22, and software stocks fell the next day as traders asked how much of the SaaS stack an AI agent can now absorb. Workday fell 9.9%, Atlassian 11.8%, HubSpot 12.7%, Salesforce 7.7%, and Okta 4.2% from Wednesday morning after the reveal, while TD Cowen analysts said the announcement was likely a major reason for a 3% selloff in the IGV software index. Presence is a managed product for putting voice and chat agents into production across support, outbound sales, billing, insurance claims, and internal IT requests. For SaaS vendors, that looks like margin compression: when a single agent platform can resolve 75% of calls and cut human handoffs by double-digit percentages in days, every subscription tied to queues and seats becomes easier to cancel. BBVA, SoftBank, and IAG are already exploring deployments in their customer workflows, exactly where traditional software vendors once claimed durable ownership. The message is blunt: owning the AI agent that runs the workflow may matter more than owning the SaaS record system beneath it.
