The supplier form is live.
The job is not.
That distinction is the whole play in Effingham County this week.
The power deal is real. The build package is not.
On July 22, OpenAI announced Project Camellia, a planned data-center campus in Effingham County, Georgia. County officials describe the proposal as a $20 billion private-capital investment inside Savannah Gateway Industrial Hub. OpenAI says it plans 3.2 gigawatts of power, delivered in phases from 2028 through 2032. Georgia Power says it signed a 25-year arrangement that gives it access to as much as 1,000 MW of flexible demand response during high demand, while Project Camellia says the final large-load contract remains subject to Georgia PSC approval. That is a serious power-and-money signal, not a normal “someday maybe” press release. It is also why a capable Southeast Georgia business should have a real plan before the invitations start moving.
Here is the cold shower: OpenAI says significant work remains on the site plan, infrastructure, phasing, design, permitting, financing, and operating model. Project Camellia says it is still running a competitive process to select the partner that will construct and operate the campus. Initial capacity is only contemplated for 2028, and the timeline still depends on approvals and development milestones. No construction partner has been named. No trade package has been released. No purchase order is hiding inside a registration form.
Do not quote a ghost job.
Get into the market before the job becomes a crowd.
The supplier list is the early door, not a purchase order
Project Camellia has already put up a local-supplier registration form for Southeast Georgia businesses. It explicitly lists construction materials, courier and delivery, environmental services, health and safety, storage, transportation, maintenance, and general support among the business types it wants to hear from. Its wording matters: register interest, then wait for an outreach-team follow-up “when appropriate.” That is a visibility move, not a bid solicitation, prequalification approval, local set-aside, scope release, work award, or invitation to price a scope. The winning posture right now is simple: be easy to understand, easy to verify, and impossible to confuse with someone chasing a headline.
OPS wrote yesterday about turbines physically arriving for Fermi’s Texas power build. That was an equipment-and-named-work signal. Camellia is different: the money and power commitment are public, but the campus team, packages, and schedule are still forming. These stories share data-center gravity; they do not share a procurement stage. The useful move in Georgia is early market entry, not a pretend estimate.
What a Southeast Georgia business should do this week
- Register only if you have a real fit. Use the local-supplier form with a direct description of what you provide, where you operate, and which jobs you can actually staff or supply.
- Build a one-page capability brief. Keep it practical: relevant industrial or infrastructure work, service radius, safety and insurance information, current capacity, lead times, and one person who can answer a serious follow-up.
- Separate market development from estimating. Do not reserve labor, buy material, offer unusual credit, or promise a client that this work is coming. There is no released scope to price.
- Watch the real gates. A named construction-and-operations partner, PSC review of the final large-load contract, permits, and a phase-one scope are the next facts that can turn interest into a bid path.
- Keep your local story ready. This project is explicitly looking at Southeast Georgia participation. Be ready to explain your geographic coverage, workforce base, and what you can deliver without stretching the truth.
Three things OPS is watching next
- High confidence: vendor registration will come before public packages. The form and the still-open partner selection process make that the most defensible near-term sequence.
- High confidence: the named construction-and-operations partner, final PSC action, and first disclosed phase will matter more than another headline about the $20 billion figure. Those are the events that create an actual buying path.
- Medium confidence: the project’s flexible-load design will make power, reliability, and infrastructure capability more visible than at a generic warehouse build. That is an inference from the 1,000 MW demand-response commitment, not a forecast of a released scope.
The smart move is not to put 2028 revenue in the forecast. It is to make sure the people building the 2028 forecast can find you before the field gets noisy.
