The Operator Model: One Person, AI, and Production-Scale Output

The Setup

The conventional technology model for a business of meaningful scope requires a team. Frontend engineer, backend engineer, DevOps, QA, project manager, technical architect. The math is well-established: a mid-market US engineering team of 4–6 developers plus support runs $960,000–$1,440,000 per year. That cost structures every decision downstream — which features to build, which markets to enter, which products to test.

The operator model proposes a structural alternative: one person, AI-enabled, building and maintaining the same infrastructure at a fraction of that cost. Not as a proof of concept. As a documented, audited, operational reality.

The data: Michael George Keating, operating with zero prior software engineering experience, built 10 production systems across 596,903 lines of code, 7 verticals, and 2 geographies in 116 calendar days. Total build cost: $67,895. Market replacement value: $795,000–$2.9 million. Monthly operating cost at steady state: $825. ROI on direct support investment: 23.1x to 84.1x. EBITDA margin moved from -$57,050/month (February 2024) to +$2,676/month (July 2025) — a $59,726 margin swing driven not by revenue growth but by infrastructure replacement.

This cluster presents the full financial picture: 28 months of QuickBooks-verified P&L, complete with the losses, the trough, the $97.7% COGS problem in affiliate marketing, and the $139,100 in outstanding liabilities. The operator model is not presented as risk-free. It is presented as a fully documented operating reality — what the numbers actually are, what they reveal about the model's strengths and limitations, and what the trajectory indicates about where it goes from here.

---

What the Data Shows

The 28-Month P&L

The aggregate 28-month financial summary (February 2024 through January 2026):

Line Item	Amount
Net Revenue	$868,147
COGS (Affiliate Payouts)	$848,031
Gross Profit	$20,116 (2.3%)
Total Operating Expenses	$166,899
EBITDA	-$146,783
Total Outside Capital Deployed	$100,899 (debt only, zero equity sold)
Monthly Operating Cost (current)	$825

The 2.3% gross margin demands context. The majority of revenue ($652,495 net) came from PRJ-12, a supplement e-commerce operation where COGS — affiliate commissions — consumed 105.9% of net revenue during the acquisition-heavy scaling phase. Strip that vertical out and the remaining portfolio runs at margins between 23% (South Africa insurance operations) and 74.8% (PRJ-16, an owned-traffic model).

The aggregate EBITDA of -$146,783 also contained a $416,552 double-counting error in affiliate payouts that appeared in both COGS and project-specific line items. Once corrected, EBITDA is -$146,783 (versus the erroneous -$563K). Financial discipline required catching that error.

The Margin Trajectory

Aggregate EBITDA is the number every evaluator asks for first. It is also the number most likely to mislead during a structural transition. The trajectory — the month-by-month margin vector — is what matters.

Phase 1 — High Revenue, Deep Losses (Feb–Apr 2024): The three highest-revenue months were also the three worst EBITDA months. February 2024: $311,443 revenue, -$57,050 EBITDA. The operation was destroying approximately $0.17 for every dollar of revenue at peak scale.

Phase 2 — The Trough (May 2024–Jun 2025): Revenue fell below $30K/month for 14 months. Average monthly EBITDA: approximately -$2,400. During this period, PRJ-01 was being built — the platform that would eliminate all SaaS and contractor costs.

Phase 3 — Margin Inflection (Jul 2025 onward): July 2025 marked the EBITDA inflection: +$2,676 on $9,673 in revenue. The mechanism was not revenue growth. It was infrastructure replacement: monthly operating costs fell from $8,367 to $825 as contractor and vendor dependencies reached zero.

The $59,726 swing from -$57K to +$2.7K monthly EBITDA came from cost elimination, not revenue addition. That is the operator model's fundamental margin lever.

Revenue Distribution

$868,147 net revenue across 18 business lines over 28 months. 75.2% concentrated in PRJ-12. 12 pre-revenue systems built at $65,394 total carry a market replacement value of approximately $1.4M–$2.9M. The question for decision-makers is not why 75% of revenue is concentrated — it is what the marginal cost of activating the 12 pre-revenue systems looks like. Answer: approximately $825/month in ongoing operating cost, because the infrastructure is already built.

The 97.7% COGS Problem

Affiliate marketing at scale produces a structural COGS problem: you pay publishers to drive customers before you know whether those customers will rebill. PRJ-12 ran at 97.7% COGS (affiliate payouts as a percentage of revenue) during the high-growth phase. The same infrastructure, shifted to owned traffic sources, runs at fundamentally different margins. PRJ-16 demonstrated this at 74.8% gross margin. The problem is not the infrastructure — it is the traffic source model.

Resilience: 99.9% Revenue Collapse and Recovery

The single highest-risk event in the portfolio: affiliate partners churned, collapsing revenue from $341,000/month to $202/month — a 99.9% revenue reduction. The infrastructure held. The operator did not exit. Monthly operating costs had already been reduced to near-zero, which meant the cost of surviving the collapse was manageable. The systems continued operating. The foundation continued compounding. The business did not cease to exist because it no longer depended on the external parties that left.

---

How It Works

The operator model produces its economics through three compounding mechanisms.

Cost structure compression. The pre-AI contractor model cost $65,054 in external development support across 3,468 delegated hours. The AI-enabled model cost $67,895 total for 10x the output at 100x the scale. Monthly operating costs fell 90% in four months and have not recovered. The cost structure does not scale with project count or system complexity — it stays at $825/month regardless of how many systems are running.

Portfolio compounding. Each system feeds the Foundation that makes the next system cheaper, faster, and higher quality. Build time fell from 24 days to 5 days. External support cost fell from $7,995 to $0. Quality improved from 31% rework (early, complex builds) to 3.7% rework (mature scaffold deployments) versus a portfolio average of 23.7% — the gap between these two numbers is the difference between operating with and without the full CEM framework. The $34,473 in direct support investment that produced the capability transition generated a 23.1x–84.1x return against market replacement value. The portfolio does not depreciate — it appreciates as Foundation depth increases.

Margin structure shift. The operator model's margin lever is cost elimination, not revenue growth. The monthly EBITDA swing from -$57K to +$2.7K was driven entirely by removing $8,367/month in operating costs — contractors, SaaS vendors, and integration overhead. AI tools at steady state: $105/month. Each cost permanently removed compounds with every subsequent month at the new baseline. Revenue can contract sharply (the portfolio survived a 99.9% revenue collapse) because the cost floor is $825/month, not $8,367/month. That asymmetry — high-ceiling revenue potential against a near-zero cost floor — is the structural characteristic that makes the operator model viable where a team model would not survive the same revenue event.

Bus factor — the honest answer. Key-person risk is elevated in a solo model. The mitigation is specific: standard frameworks (PHP/Laravel), full version control across 2,561 commits, production documentation. Any competent mid-level developer can read, maintain, and extend the codebase. GitHub's 2024 Octoverse Report finds that 73% of open-source projects have a bus factor of one or two — and the vast majority continue through contributor transitions. The relevant question is not whether one person built it, but whether the system is maintainable by others. It is.

The risks the model eliminates — vendor lock-in, contractor dependency, coordination overhead, integration fragility — are structural and ongoing. The risk it introduces — key-person concentration — is bounded and mitigable.

---

The Articles

$67,895 vs $2.9M: The New Build-vs-Buy Math for Private Equity Portfolio Companies — The cost comparison with external benchmarks and the compounding advantage that makes the gap widen over time.

How to Move from Contractor Dependency to In-House AI-Enabled Development — The four-phase transition from 69% external dependency to 100% solo execution, with the methodology at each stage.

Software Quality at Scale: How to Achieve a 12.1% Defect Rate vs the 20-50% Industry Average — The quality controls that produced a 12.1% product bug rate while maintaining 4.6x velocity increase.

The Compounding Software Portfolio: How 10 Systems Each Made the Next One Cheaper to Build — The three compounding curves: build time, cost, and quality — all moving simultaneously in the right direction across 10 systems.

Bus Factor Risk in Solo-Built Software: How to Ensure Continuity — The honest risk assessment, the mitigation framework, and why key-person risk ranks below revenue concentration and platform dependency in most operators' actual risk hierarchy.

116 Days of Sustained Development Output Without Burnout: The Evidence — The commit-level evidence for sustained output: 13.4x output multiplier with zero burnout incidents across the full window.

28 Months of P&L Data: The Full Financial Picture of an AI-Enabled Operation — Complete unaudited-to-QB-verified financial history: revenue, COGS, EBITDA by month, the double-counting error, and the current position.

How to Survive a 99.9% Revenue Collapse: Infrastructure Resilience for Operators — The affiliate churn event: $341K to $202/month in revenue, what held, what broke, and what made survival possible.

Complete Product Lifecycle: From $0 to $173K/Month to Controlled Wind-Down — The full arc of a DTC product: launch, scale, peak, margin compression, and managed wind-down — with financial data at each stage.

38 Products Tested, 6 Scaled: Portfolio Discipline and the Power Law in Practice — The test machine: why 32 failures combined cost less than one traditional product launch, and what the 6 winners produced.

Revenue Distribution Across 18 Business Lines: Where $868K Went — The full 28-month revenue map: concentration, margin by line, and what the pre-revenue systems represent in optionality.

The 97.7% COGS Problem in Affiliate Marketing (and How to Fix It) — How affiliate COGS consumed 105.9% of revenue at peak scale, what caused it, and the structural shift to owned traffic that changes the margin profile.

From -$57K to +$2.6K Monthly EBITDA: The Margin Trajectory of an AI-Enabled Operation — The three phases of the margin journey, the mechanism behind the inflection, and what the trajectory indicates.

15 Attribution Views Across 7 Dimensions: How to Build Real Revenue Visibility — The analytics infrastructure behind 18 business lines: how to understand which revenue is real and which is masking structural loss.

---

Frequently Asked Questions

Can One AI-Enabled Operator Really Replace a Full Development Team? — One operator replaced $960K/year in engineering team cost for $67,895 total — here is exactly what that required and what it produced.

What Is the ROI of an AI-Enabled Operator Compared to a Traditional Engineering Team? — 23.1x–84.1x on audited investment, with the compounding advantage that makes the gap widen every subsequent month.

How Do You Evaluate the Capability of an AI-Enabled Operator? — The metrics that matter: commit velocity, rework rate, defect category distribution, system complexity, and portfolio depth — not LOC or years of experience.

What Are the Risks of Relying on a Solo Operator for Technology? — Bus factor risk is real and bounded — here is the honest risk hierarchy and why key-person risk ranks below revenue concentration and platform dependency.

How Does the AI-Enabled Operator Model Scale Beyond One Person? — The Foundation infrastructure scales with additional operators — here is what adding a second person to a mature CEM system looks like.

What Does a Fractional Operator Engagement Look Like? — The scope, deliverables, and cost model for a fractional engagement versus a full-build operator model.

What Does the Full P&L Look Like for a Solo AI-Operated Business? — 28 months of QB-verified financials: $868K revenue, -$147K EBITDA, $825/month current operating cost, $139K in outstanding liabilities. The full picture.

How Do You Recover From a Major Revenue Loss as a Solo Operator? — The 99.9% revenue collapse: what the infrastructure that survived it looked like, and what the recovery required.

---

What This Means for PE Firms and Executive Decision-Makers

The operator model is a financial thesis as much as a technology thesis. The claim is that one AI-enabled operator, with the right methodology and infrastructure, can produce output that historically required a team — at 2–8% of team cost, with compounding returns on every subsequent project, and zero equity dilution.

The 28-month P&L presented in this cluster is not a highlights reel. It includes the -$57K month, the 14-month trough, the 97.7% COGS failure, and the $139,100 in outstanding liabilities. The transparency is deliberate. Credible operator evaluation requires the full picture — not a filtered version.

What the full picture shows: a structural transition from rented infrastructure to owned infrastructure, from external dependency to internal capability, from linear cost scaling to compounding cost reduction. The EBITDA inflection in July 2025 was not driven by a revenue event. It was driven by the completion of the infrastructure build that eliminated the cost structure that had been producing the losses.

That is the operator model's core proposition: build the infrastructure once, eliminate the recurring cost permanently, and let the compounding advantage separate you from everyone still renting.

---

References

1. Morgan Stanley (2024). "Private Market Valuation Analysis." Institutional research on PE portfolio company valuations and margin trajectory benchmarks.
2. Bain & Company (2024). "Global Private Equity Report." Annual analysis of PE hold-period performance and value creation levers.
3. McKinsey & Company (2024). "Digital Operations Research." Research on technology stack consolidation and AI-enabled operational efficiency.
4. Gartner (2024). "IT Spending Forecast." Enterprise software spending projections and SaaS cost escalation trends.
5. GitHub (2024). "Octoverse Report." Annual analysis of open-source development patterns, including bus factor distribution across projects.
6. Glassdoor/Levels.fyi (2024-2025). Compensation data for mid-market US engineering teams.
7. CB Insights (2024). "State of Venture Report." Startup failure rates and capital allocation analysis.
8. Harvard Business School (2023). "Startup Failure Research." Analysis of venture-backed startup outcomes and capital return rates.
9. PitchBook (2024). "Middle-Market PE Exit Analysis." Revenue concentration and business line diversification benchmarks for PE portfolio companies.
10. Keating, M.G. (2026). "Case Study: The Full Portfolio." Stealth Labz. Read case study
11. Keating, M.G. (2026). "Case Study: The Cost Inversion." Stealth Labz. Read case study
12. Keating, M.G. (2026). "Case Study: The Platform Displacement." Stealth Labz. Read case study