OpenAI's Presence resolves 75% of support calls without a human. Trade contractors already have a cheaper version.
OpenAI launched Presence, an enterprise AI agent platform that now handles three-quarters of its own support calls without a person — but it's priced and deployed like consulting, not software. Mechanical, electrical, and plumbing subs already run cheaper AI dispatch tools built on the same underlying models.
OpenAI's newest product isn't a chatbot upgrade — it's a claim that AI can now run a company's phone line. Presence, which OpenAI launched on July 22, resolves 75% of inbound calls on its own English-language support line without a human stepping in, according to the company. For a mechanical, electrical, or plumbing sub that lives and dies by its dispatch and service-call line, that's a real number worth paying attention to. The catch: Presence isn't for sale to a 50-person trade contractor, and a cheaper version of the same idea is already running on their phones.
What did OpenAI actually launch?
Presence is a platform for building voice and chat agents that combine a company's own policies, guardrails, and escalation rules with OpenAI's models, so the agent can answer questions, pull account context, take pre-approved actions, and hand off to a person when a rule says it should. OpenAI is running it on its own support line — reachable at 1-888-GPT-0090 — where it says the system verifies callers, resolves issues, and now handles 75% of calls end-to-end. A Codex-powered feedback loop reviews the agent's real conversations and proposes behavior changes that staff test and approve, which OpenAI says cut human handoffs by 15 percentage points in 10 days. Early enterprise use cases are customer support, sales, and internal workflows at banks, telecoms, and insurers.
Why doesn't this reach a trade contractor yet?
Presence isn't a product you sign up for. It rolled out through limited general availability, backed by OpenAI's $10 billion Deployment Company (formed after its acquisition of the consultancy Tomoro), and deployments are led by OpenAI's own Forward Deployed Engineers or select systems integrators. Pricing isn't public — it's scoped per customer, which multiple outlets have flagged as enterprise-consulting rates rather than a monthly software bill. That model fits a call center handling tens of thousands of contacts a month. It doesn't fit a mechanical sub's two-person dispatch desk.
What does a trade contractor actually have access to right now?
A market of AI phone-answering tools built specifically for contractors already exists, and several run on the same underlying models Presence uses. Services like OnCrew and Bland AI sell AI voice agents that answer, qualify, and route service calls for HVAC, plumbing, and electrical companies today, priced closer to the traditional live-answering-service range than to enterprise consulting.
| OpenAI Presence | Contractor-market AI answering (OnCrew, Bland AI, similar) | Traditional live answering service | |
|---|---|---|---|
| Who it's sold to | Large enterprises (banking, telecom, insurance) | Trades and home-service contractors | Trades and home-service contractors |
| Deployment | OpenAI Forward Deployed Engineers / systems integrators | Self-serve signup | Self-serve signup |
| Pricing | Scoped per customer, not public | Roughly $50–$300/month depending on volume | Roughly $400–$800/month for full 24/7 coverage |
| Escalation/guardrail layer | Built and tuned by OpenAI + customer policy team | Vendor-defined, often less customizable | Human dispatcher judgment |
What should a contractor evaluating any of these actually check?
Presence's real contribution isn't the phone call itself — voice AI answering trade calls isn't new. It's the explicit policy and escalation layer: a written rule set for what the agent can decide alone and what it must hand to a person, tested before it goes live and revised as those rules change. That's the question worth asking any AI answering vendor, not just OpenAI's: can you see the escalation rules in writing, and do they name every life-safety call type — gas odor, no heat in freezing weather, active flooding, electrical burning smell — as an automatic route to a human, every time? A tool that can't answer that isn't ready for a service line, regardless of price.
Construction AI Brief tracks which AI-industry launches actually reach a trade contractor's back office — new pieces most days at constructionaibrief.com.
- What is OpenAI Presence?
- Presence is an enterprise AI agent platform OpenAI launched on July 22, 2026, for building and running voice and chat agents that answer questions, use company systems, take approved actions, and escalate to a person when a policy requires it. OpenAI says it now resolves 75% of inbound calls on its own English-language support line without human help.
- Can a mechanical or HVAC contractor use OpenAI Presence for its dispatch line?
- Not in the way it's currently sold. Presence is a limited general-availability rollout deployed through OpenAI's Forward Deployed Engineers and select systems integrators at scoped, enterprise-consulting rates — not a self-serve product a 50-person sub can sign up for. It's built for high-volume call centers in banking, telecom, and insurance first.
- How much does OpenAI Presence cost?
- OpenAI hasn't published pricing. Each deployment during the limited GA phase is scoped individually to the customer's use case, which multiple outlets have described as enterprise consulting rates rather than a subscription price.
- What's the difference between OpenAI Presence and AI answering services built for contractors?
- Contractor-focused AI answering tools (OnCrew, Bland AI, and similar) already offer self-serve, monthly-priced AI phone agents built on top of models from labs including OpenAI. Presence is OpenAI's own attempt to package the guardrail, escalation, and continuous-improvement layer around those models — the part that decides when the AI should stop and hand off to a person.
- What should a trade contractor's after-hours line still send to a human?
- Anything with a life-safety component — gas smell, active water intrusion into occupied space, no-heat calls in freezing weather, electrical arcing or burning smell — should route to an on-call tech immediately, not sit in an AI triage queue. That's exactly the kind of policy-defined escalation rule Presence is designed to enforce, and any AI answering tool a contractor evaluates should be able to show the same rule in writing.