Cost analysis
One operator with AI tooling, against a modelled conventional team, over 117 days of one real project.
The two columns
Scope: Campaign Brain Last 3 Months.
| Team size | 1 |
| Commits | 27,939 |
| Repositories | 134 |
| Issues opened | 4,601 |
| Issues closed | 3,096 |
| Operator cost | $60,000 |
| AI cost | $800 |
| Total | $60,800 |
The modelled team
10 people at $1,806,000 a year fully loaded. Salary figures are market rate; the 1.4x loading covers benefits, tax and overhead.
lower costhigher
The gap
95.3% lower cost over the window, and 9 months to 1 daysin elapsed time. Both figures compare a measured column against a modelled one and inherit that model’s assumptions entirely.
The velocity multiplier the pipeline computes is 1,710x. It is not quoted as a headline here, because it divides by an active-day count of 1, and a ratio with a denominator that small is arithmetic rather than evidence.
What the work covered
What the research says
Published findings on AI-assisted development, for calibration against the single project above.
| Source | Finding |
|---|---|
| GitHub/Microsoft 2022 | 55% faster task completion |
| McKinsey 2023 | 20-45% productivity improvement |
| Google 2024 | 25%+ of new code AI-generated |
| BCG/Harvard 2023 | 40% higher quality output |
| Deloitte 2024 | 25-35% project cost savings |
cost-model.json, generated Sep 6, 2026