$206,000 is the number almost nobody puts in the hiring proposal: base salary, recruiter fees, and the months of ramp-up before that person is productive. And that is just the first hire.
The breakdown the job offer does not show
- Senior AI salary in the US: the basis of the number, but not the total.
- Recruiter fees: 15% to 25% of annual salary, paid once — but on every hire.
- Ramp-up: three months is typical before the engineer delivers at full speed, with the existing team absorbing that context cost.
- Bus factor: if that person leaves at month ten, the cost repeats.
Traditional outsourcing does not solve this, it relocates it
Cutting the hourly rate by hiring offshore sounds attractive until the problem stops being price and becomes lost context: the ticket describes the symptom, not the system, and quality turns variable because nobody on the other side understands why that architectural decision was made six months ago.
What changes with an embedded squad
A forward deployed squad costs more per person per hour than a traditional hire. But the context ramp-up does not repeat per person, because the squad already operates with a squad lead, seniors and dedicated QA from day one — not as roles that get filled in over time.
The question worth asking is not what the hour costs. It is what the time between “we have the problem” and “it is in production and it held” costs.