Engagements & pricing
Planning ranges, not one-size-fits-all packages.
BrooksFlow prices private AI work based on scope, data sensitivity, deployment model, hardware requirements, integrations, on-site needs, and operating cadence. The ranges below are planning ranges — every engagement is quoted against its actual scope.
Planning ranges
What engagements typically cost.
Private AI Readiness Assessment
Typically $7,500–$15,000
Roadmap, risk review, hardware recommendation, and pilot proposal. Most delivered in 5–10 business days.
Private AI Pilot
Typically $35,000–$95,000+ implementation
Plus customer-paid hardware and third-party costs. One contained workflow with real users.
Department Deployment
Typically $100,000–$250,000+ implementation
Plus customer-paid hardware and third-party costs. Proven workflows expanded across a department.
Private AI Room / Rack
Custom · typically $250,000+ total program
Depends on hardware, facility, security, and support requirements. Scoped as a program, not a package.
Managed AI Ops
Typically $3,500–$25,000+/month
Depends on workflows, users, support cadence, and security requirements.
Ranges are for planning. Final quotes follow scoping — and the assessment exists so neither side is guessing.
Hardware payment
Hardware is paid by the customer, upfront.
Hardware, GPUs, servers, racks, storage, networking, power, cooling, shipping, taxes, third-party software, and facility work are paid separately by the customer. BrooksFlow can recommend, configure, and manage procurement, but does not finance customer hardware. Hardware orders require customer approval and upfront payment before procurement.
What moves the number
What determines where you land in a range.
- Scope: number of workflows, users, and integrations
- Data sensitivity and the controls it requires
- Deployment model: on-premise, private cloud, or customer-controlled cloud
- Hardware requirements and facility work
- On-site needs and travel
- Operating cadence: support expectations, reporting, and review cycles
FAQ
Pricing questions
The customer, directly and upfront. BrooksFlow recommends, configures, and manages procurement, but does not finance customer hardware. Orders are placed only after customer approval and payment.
Not always. Some workflows run well on CPU or a single workstation-class GPU. The assessment sizes hardware to the actual workload — not every pilot requires a full rack or a large GPU server.
Because scope, data sensitivity, deployment model, and hardware vary enormously between organizations. Ranges keep planning honest; the assessment turns a range into a firm quote.