Pilot launch — operations
This page is the operating model behind the Academy: how twelve Socratic Guides are found and certified, where $662,900 goes across a twelve-month pilot, and the conditions under which we would stop. It exists so a parent, a prospective guide, or a sponsor can read the same numbers rather than three different pitches.
Target certified guides
12
First cohort, staged across four onboarding phases
Pilot horizon
12 months
September 2026 through August 2027
Total pilot budget
$662,900
Operating plus one-time retrofit capital, modelled monthly
Full campus build
$3M–$8M+
Land, facility, systems, and landscaping — separate from the pilot
Status label. Every figure on this page is a planning model, not a result. The budget is an assumption set we expect to revise once real invoices arrive. The candidate names in the workbook are illustrative placeholders used to test the tracker structure — no person has been recruited, screened, or offered a stipend.
Why this pilot
Dependent offloading — handing judgement to a tool because the tool answers faster than thinking does. It is cheap, it feels like competence, and it compounds. The pilot's entire method is built to make the cheap path visibly unsatisfying.
Autonomous offloading: using the machine for retrieval and arithmetic while keeping the judgement, the framing, and the burden of defending an answer with the student. This is what the AI stations are for, and it is what guides are certified on.
A teacher is measured on coverage. A guide is measured on whether the student can defend an answer to a machine that disagrees. Those two jobs reward different behaviour, so we recruit and certify for the second one directly.
The attention argument
Guide pipeline
A resting heart-rate-variability reading plus a live role-play with a frustrated learner. The question is whether the candidate stays regulated while the student is not. Candidates below the working threshold are offered coaching rather than rejection.
Scored one to five on whether the candidate asks the next question instead of delivering the next answer. The failing pattern is well-meant lecture; the passing pattern is patient inquiry with real wait time.
The candidate sits at an AI station with a student and must catch a sycophantic answer, demand a disagreement, and hand the judgement back to the learner. Scored one to five on the catch, not on the technology fluency.
Two panel members independently review recorded dialogues before certification. Split decisions return the candidate to the sandbox rather than resolving upward.
Applied
Application and background checks complete; no assessment yet.
Autonomic ER screen
In the regulation and role-play assessment.
Socratic sandbox
At an AI station with a student, being scored on the catch.
Socratic mirror
Recorded dialogues under review by two panel members.
Certified
Cleared to guide unsupervised, with quarterly re-review.
| Phase | Window | Seats | Focus |
|---|---|---|---|
| Phase 1 | September 2026 | 4 guides | Founding cohort. Literature, philosophy, early childhood, and one working musician. These four write the observation protocol the later phases inherit. |
| Phase 2 | October 2026 | 4 guides | First expansion. Pedagogy and movement specialists join; the dance-and-cognition sessions begin here, and stipend load rises from $14,000 to $21,000 per month. |
| Phase 3 | November 2026 | 4 guides | Developmental psychology, computational design, and natural science. Outdoor sessions move from occasional to scheduled, and a second recreation guide is added. |
| Phase 4 | December 2026 – January 2027 | 6 candidates in review | Deepening bench: classical philosophy, mathematics, narrative arts, engineering, and music. Not all Phase 4 candidates certify inside the pilot year, and that is expected. |
Financial model
Roughly two-thirds of the pilot is people. That ratio is the argument: the expensive part of this school is the adults in the room, not the technology. Compute and licences together account for eleven percent, and the facilities line is front-loaded into the first three months as one-time retrofit work.
What this budget does not cover. The $662,900 figure is an operating and retrofit model for a pilot year inside an existing or donated space. Buying land, constructing a purpose-built facility, hiring architects and building-technology specialists, and landscaping the grounds would require several million dollars at minimum. Corporate donations of hardware or software might reduce the visible technology line, but the expert time to design, integrate, and commission building systems, landscape, and outdoor learning infrastructure would cost many multiples of the $73,400 technology allocation. That capital phase is kept separate so the pilot can be judged on its own merits before anyone commits to a build.
Move the donation levels and watch what actually changes. In-kind hardware and software reduce the $73,400 technology allocation — eleven percent of the pilot — and nothing else. The expert time to design, integrate, and commission building systems, landscape, and outdoor learning infrastructure is a separate, larger cash line that donated equipment does not touch unless firms also donate the labour.
Corporate hardware gifts are the most commonly offered form of support.
Usually granted for a fixed term, so full donation is rarely permanent.
Planning assumption, not a quote: architects, building-technology specialists, landscape and commissioning work. Sits in the capital phase, outside the $662,900 pilot.
The only input that meaningfully moves the total. Treat high values as optimistic until signed.
Against a no-donation baseline of $367,000, this scenario removes 0.0% of the cash requirement. Expert time is 80.0% of what remains.
Set a donation level to compare scenarios. The baseline assumes the Academy pays for everything.
Falsifier. If a signed capital proposal shows expert design, integration, and commissioning costing less than twice the technology allocation, the multiplier assumption here is wrong and this calculator should be rebuilt around the real bid.
$424,500 · 64%
$130,500 · 20%
$73,400 · 11%
$34,500 · 5%
The full workbook — dashboard, guide onboarding tracker, and the month-by-month budget with live formulas — is available for sponsors and advisers who want to change the assumptions and see what breaks.
Download the operations workbook (XLSX)Site strategy
The restored home and recording studio near Nashville remain the operational base: writing, curriculum recording, guide training, advisory meetings, and planning. They are not the school site. Residential zoning, homeowners' regulations, occupancy limits, parking, fire-code, and safeguarding requirements create a heavy and unnecessary regulatory bar for a home-based school.
The $662,900 pilot budget is therefore scoped for operations and interior retrofit inside a leased, donated, or hosted non-residential space — for example, a surplus school wing, a church hall, a corporate training suite, or a community-arts building. That keeps the pilot focused on the method, not on rezoning a residence.
Read the location hypothesis →
The several-million-dollar capital phase is for a future, purpose-built campus elsewhere in the Franklin/Brentwood/Nashville area. The home stays base operations; the school gets its own compliant facility when the method is proven.
Corporate development
The pilot budget assumes the money exists. It does not yet. Corporate and foundation support runs on annual cycles and formal diligence, so the honest planning figure is eighteen months from first serious conversation to usable cash — not the three or four months a launch date tends to assume. Fundraising also has its own cost, which belongs in the model rather than hidden inside someone's unpaid evenings.
| Window | Stage | What happens | Estimated cost |
|---|---|---|---|
| Months 1–3 | Prepare the ask | Audited-quality financial model, one-page brief, sponsor deck, and a legal/fiscal home — either a fiscal sponsor or a filed nonprofit. Nothing is sent to a company until these exist; a first meeting spent apologising for missing paperwork rarely gets a second.Formation and filing, accountant review, design and copy for the brief pack. | $18,000–$34,000 |
| Months 3–7 | Warm-list outreach | Roughly 60–120 qualified targets: regional employers, technology firms with education-giving lines, community foundations, and family offices. Expect a 10–20% first-meeting rate from warm introductions and low single digits from cold approaches.Part-time development contractor, CRM, travel to regional meetings. | $12,000–$28,000 |
| Months 6–12 | Diligence and structuring | Serious corporate funders run diligence: governance, insurance, child-safety policy, data handling for the AI stations, and outcome measurement. This is the stage that most often stalls a first campaign, because the answers cannot be improvised.Policy drafting, insurance binders, safeguarding and data-protection review, evaluation design. | $20,000–$45,000 |
| Months 9–18 | First closes | Corporate cycles are annual. A pledge made in month ten commonly funds in the next fiscal year, so plan for a 12–18 month gap between the first serious conversation and usable cash. Anchor gifts arrive before the long tail, not after.Grant writing, reporting setup, board and advisory-panel time. | $10,000–$22,000 |
| Months 18–30 | Capital conversations | Only once the pilot has run a term does the land-and-build conversation become credible. Capital campaigns for several-million-dollar builds typically take two to four years of their own and are staffed differently from operating fundraising.Capital-phase development is a separate budget with its own staffing. | Not modelled here |
$60,000–$129,000 · 9–19%
Expressed as a share of the $662,900 pilot. A first campaign at this scale commonly costs ten to twenty cents per dollar raised; established campaigns run lower. Treat any figure under ten percent as a claim that needs evidence.
Interim capacity
Between funding and opening there is a stretch of unglamorous work: filings, policy, procurement, receiving deliveries, coordinating trades. Hiring permanent staff for it is expensive and hard to unwind. These are the routes we would use instead, with their real costs rather than the assumption that volunteers will absorb it.
An established nonprofit holds the funds and provides bookkeeping, payroll, and insurance for a fee of roughly 5–10% of funds raised. Slowest to feel like your own organisation, fastest to become fundable.
Cost: 5–10% of raised funds. Benefit: no HR build-out in year one.
A part-time grant writer and a fractional bookkeeper, engaged monthly rather than hired. Typical range is $45–$95 per hour for development work and $40–$70 for bookkeeping in most regional markets.
Cost: predictable monthly retainer. Risk: divided attention across clients.
For the retrofit window — receiving deliveries, inventorying equipment, coordinating trades — agency temps at a 25–45% markup over wage are cheaper than a permanent hire for a three-month burst.
Cost: markup on wage. Benefit: no severance or ramp-down problem.
One experienced construction or facilities coordinator, engaged part-time, to hold trades and vendors accountable during retrofit. Commonly 2–5% of construction value, or a day rate for smaller scopes.
Cost: real, and usually recovered in avoided change orders.
Co-working desks, a sponsor's spare suite, or space inside a partner school or church. Assume 6–18 months of donated space is possible; assume nothing beyond that without a signed agreement.
Cost: near zero. Risk: a landlord's change of plan resets your timeline.
Many companies fund employee pro bono time for finance, legal, IT, and project management. This is the one form of in-kind support that reduces expert-time cost rather than only hardware cost.
Cost: coordination time. Benefit: the scarce input, not the cheap one.
Practicum and internship placements in education, social work, and facilities management supply supervised hours at low cost. They require real supervision, so they are additive capacity, not free labour.
Cost: supervisor time. Benefit: recruitment pipeline for later hires.
The same multigenerational pool that supplies guides includes retired administrators, controllers, and project managers. Some will take an operations role for stipend or expenses only.
Cost: low. Constraint: availability is seasonal and personal.
The same caution as the donation calculator. Donated desks, temps, and interns lower the cheap lines. Skills-based corporate volunteering is the only route above that reduces the expensive line — expert professional time — and it should be asked for by name rather than hoped for as a by-product of a hardware gift.
Falsifier ledger
Each claim below carries the observation that would retire it. A pilot that cannot be falsified is not a pilot; it is a promotion.
Emotionally regulated retirees make better Socratic guides than newly trained young graduates.
Falsifier — If certified guides drawn from the retiree pipeline show no advantage over a matched group of early-career guides on student explanation quality after two terms, the recruitment thesis is wrong and the pipeline should be opened by aptitude rather than by life stage.
The desk-free, biophilic room improves sustained attention.
Falsifier — If sustained-attention measures in the retrofitted room match a conventional classroom running the same Socratic curriculum, the $130,500 facilities line is decoration and should be cut, not defended.
AI stations used as an adversary, not an oracle, reduce dependent offloading.
Falsifier — If students at the stations increasingly accept first answers without demanding a disagreement across the pilot year, the Explorer-in-Charge protocol has failed and the stations should be removed until it is redesigned.
The pilot budget is sufficient to certify twelve guides in twelve months.
Falsifier — If fewer than eight guides certify by month twelve at the modelled spend, the cost per certified guide is understated and the model needs rebuilding before any sponsor is asked for a second year.
A full campus can be delivered within the pilot's capital envelope.
Falsifier — If land acquisition, construction, building-technology systems, and landscaping cannot be realistically scoped in the several-million-dollar range, the pilot is not a path to a standalone campus and must be reframed as a hosted or leased-site program until capital is secured.
The founder's home can serve as the pilot school site.
Falsifier — If residential zoning, homeowners' association rules, fire-code occupancy limits, parking, ADA access, or safeguarding inspections block the use of the home as a school facility, the pilot must operate from a leased, donated, or hosted non-residential site. The home remains base operations only.
Corporate and foundation support for the pilot year can be raised in roughly eighteen months.
Falsifier — If no anchor commitment covering at least a quarter of the year is signed within eighteen months of the first qualified meeting, the sponsor thesis is wrong for this market and the pilot must shrink to a tuition-and-family-funded single cohort or wait.
Temporary, fractional, and volunteer capacity is enough to carry the pre-opening build-out.
Falsifier — If the retrofit slips more than one quarter, or if filings and safeguarding policy are still incomplete at the point families are ready to enrol, the interim staffing model has failed and a paid full-time operations hire must be added to the budget before opening.
Next