portfolio/.project/VISION.md

Autonomous Enterprise Portfolio

Vision

Document: VISION.md
Status: Founding vision — initial version
Established: September 16, 2026
Owner: Ken Schaefer

Build once. Launch rapidly. Validate with a sale. Operate, replicate, or sell.

1. The vision

Create a diversified portfolio of complementary, AI-enabled businesses that can be conceived, launched, commercially tested, and operated with dramatically less incremental capital and routine human labor than conventional small businesses require.

The portfolio will use shared infrastructure, reusable business capabilities, and independently separable ventures. The owner will remain actively involved in vision, discovery, product development, editorial judgment, relationships, and major decisions, while AI and automation take over as much repetitive operational work as they can reliably perform.

Every venture is an experiment with a defined validation window. Ventures that establish commercial demand may become ongoing income sources, grow into stand-alone businesses, be licensed, or be sold. Unsuccessful experiments will be closed responsibly; their reusable assets and lessons will return to the portfolio.

The long-term ambition is not merely to own several businesses. It is to develop a repeatable capability for creating and testing businesses.

2. Why we are doing this

The initiating challenge came from videos claiming that AI can build highly profitable businesses in extraordinarily short periods, including a claim of a $1.2 million business built in roughly 30 minutes. The revenue, profit, timing, and repeatability of those claims have not been established. Building a website or product quickly is not the same as earning its claimed revenue.

We will neither accept such claims at face value nor dismiss their underlying possibility. We will test the proposition ourselves, recording build time, time to first sale, actual collected revenue, incremental spending, human effort, and failure as carefully as success.

The broader hypothesis is that AI may lower the minimum viable scale of a business: one technically capable person may be able to undertake work that previously demanded a team, agencies, or extensive founder labor. At the same time, cheaper production can increase competition, making demand, distribution, trust, and differentiation no less important. The experiment must test commercial outcomes, not simply prove that we can generate outputs.

3. The hypotheses we will test

1. Launch compression: Can AI and existing assets shorten the time from idea to purchasable product to days or weeks?
2. First-sale compression: Can a newly launched venture obtain its first genuine paying customer within a bounded validation window?
3. Capital efficiency: Can we run many experiments at little incremental cost by reusing infrastructure and capabilities already under our control?
4. Operational leverage: Can shared AI systems carry out opportunity discovery, sales support, production, administration, and routine operations without requiring proportionate increases in the owner's labor?
5. Commercial adjacency: Can an audience, product, service, or discovery in one venture create customers or useful intellectual property for another?
6. Transferable value: Can validated ventures be sufficiently modular to operate independently, be licensed, or be sold?
7. Factory repeatability: Can a repeatable process support numerous experiments rather than requiring each business to be invented from scratch?

These are questions to investigate, not claims of proven capability.

4. Portfolio philosophy

4.1 Embrace uncertainty, bound exposure

We are deliberately willing to go out on a limb. An initial 70% allocation of available effort to speculative opportunities is an expression of that ambition, not a promise to neglect near-term financial commitments. Actual allocation will be set and rebalanced in the portfolio plan as cash and evidence change.

We distinguish uncertainty about demand, timing, and outcomes from exposure to cash loss, contractual commitments, and other harm. Experiments should be bold in their hypotheses and disciplined in their commitments.

4.2 First sale is the first commercial gate

The initial commercial KPI for each new venture is one genuine, arm's-length sale with collected revenue greater than $0 within its defined validation window. A sale earns the venture consideration for further investment; it does not prove profitability, repeatability, or sustainability.

A venture that records no qualifying sale by its deadline fails its current commercial hypothesis: we stop investing under that hypothesis and decide whether to retire, archive, or explicitly redesign the experiment. We do not quietly extend deadlines or mistake activity for revenue.

Pledges, expressions of interest, waitlists, booked but unpaid orders, and promotional exchanges are evidence of different kinds—not interchangeable with collected sales. Crowdfunding proceeds, preorders, subscriptions, advertising receipts, and product purchases must be identified by type; money subject to fulfillment, refund, or other obligations is not automatically profit.

Existing revenue relationships are evaluated incrementally. Gaussian, for example, has already produced more than $8,000 in 2026; the new test concerns additional paid work, not proving the first-ever sale again.

4.3 Keep ambition separate from the acceptance test

Our provocative stretch outcome is to launch a business in seven days and collect $10,000 in gross revenue during its first 30 days of operation. It is a challenge to test, not a forecast and not a prerequisite for passing the first-sale gate.

We will separately record gross receipts, refunds and obligations, variable and incremental costs, profit or loss, and required human effort. A business with its first small sale may merit further investigation; a business with large gross revenue but unsustainable economics may not.

4.4 Validate before perfecting

A purchasable, useful initial product can precede a polished platform and a fully autonomous operation. We favor minimum viable offers, paid pilots, limited releases, and carefully scoped Early Access where suitable. The product must deliver present value, and its limitations and discontinuation terms must be candid. Early Access does not erase customer, security, privacy, platform, or legal obligations.

4.5 Complementary businesses, not unrelated bets

We will seek combinations in which content builds an audience for products; products create material for content; services reveal software opportunities; software and workflows become reusable intellectual property. We will still test each connection rather than assume that an audience will buy or that one revenue stream guarantees another.

The portfolio must also retain genuine diversification: sharing infrastructure is useful, but shared distribution or a single dependent customer can create correlated failure.

5. The business factory

The enterprise will have shared services and separable business units.

Shared capabilities may include opportunity intelligence, market research, sales development, marketing, content production, AI inference, engineering, accounting support, reporting, and deployment. They will be implemented with an appropriate combination of specialized agents, deterministic workflows, existing software, and people. We will not presume that every business function needs its own agent.

Each venture should have an identifiable offer, audience, brand or product identity where appropriate, assets and intellectual property, customer records, financial attribution, operating configuration, validation criteria, and a path to stand-alone operation.

Modularity is a founding constraint: a venture should be possible to retain, scale, license, sell, or retire without dismantling the shared factory or unnecessarily entangling other ventures. This does not require a separate LLC, bank account, cloud account, or full software stack for every early experiment. It does require deliberate boundaries, documentation, portable data, and clarity about ownership and transferability. Third-party accounts, contracts, licenses, and customer data may not be transferable without permission or additional work.

The factory itself—its methods, workflows, software, templates, and operating know-how—may ultimately become intellectual property or a commercial offering. That possibility must not delay the first experiments.

6. The owner's role and the purpose of autonomy

The objective of automation is to remove required routine labor, not to remove the owner from work he finds valuable. Ken intends to remain involved in opportunity selection, developing products, architecture, creative direction, major customer relationships, portfolio allocation, and consequential decisions.

We will focus early automation on his constraints: finding leads, qualifying opportunities, supporting sales, following up, and covering business functions for which time is scarce. His comparative advantage is in building and developing, not spending every hour prospecting or administering the portfolio.

An Opportunity Intelligence capability should independently discover, evaluate, and organize relevant employment, contract, and client opportunities, then present actionable findings for Ken to review and act upon. Independence in research does not imply authorization to apply, contact people, accept terms, move money, or make irreversible commitments without appropriate controls.

Autonomy is progressive: prove a bounded workflow, measure quality and exceptions, expand permissions where earned, and preserve human governance of consequential decisions. We want the businesses to continue their routine work when Ken turns his attention elsewhere—not a theatrical claim of zero human involvement.

7. Assets and capital strategy

We begin with assets already under our control:

• A home lab with static IP, firewall, reverse proxy, virtualization, and a Docker server for public-facing services.
• The Forge / Gateway / Oracle AI infrastructure, local-model capability, and Project Thoth engineering work.
• Fractional Insight CIO, LLC, its registration, EIN, business bank account, QuickBooks, and credit-card processing capability.
• Professional experience, reputation, career documentation, existing consulting relationships, and defined service offerings.
• Withered Sanctum and Pyramid, including existing creative work and historical publishing experience.

Previously committed lab and business expenditures are part of our existing cost base; the investment decision for an experiment focuses on incremental spending and labor. We will still account for any additional electricity, model/API usage, storage, hardware load, fees, and support attributable to a venture. Existing resources are an advantage, not a claim of zero cost or unlimited capacity.

Lab first, commercial infrastructure when justified. Each software venture's detailed plan will model the cost of a suitable Azure, DigitalOcean, or other production deployment and calculate the revenue or contribution needed to support it. Commercial hosting cost creates an economic affordability floor, not an automatic command to migrate. Capacity, security, customer-data sensitivity, recovery needs, and promises to customers may require a different environment sooner; if so, we must fund it or narrow the offer rather than make commitments the lab cannot safely support.

Customer-funded development is an option to examine in every relevant business canvas: paid pilots, preorders, Early Access, crowdfunding, memberships, or subscriptions. We will distinguish committed funding and collected cash from earned revenue and account for fees, refund risks, and delivery obligations.

8. Initial portfolio ecosystems

A. Withered Sanctum: media, games, and publishing

An Ancient Egypt channel is the first speculative proof of concept. It may connect to Pyramid tabletop products, historical and game-related digital publications, assets, memberships, and eventually an interactive or video-game adaptation. The channel is a potential audience and distribution asset, not necessarily an advertising-only business.

The initial aggressive seven-day goal is to launch a real, original video and a repeatable production process. That establishes operational feasibility; the commercial gate is the first attributable paid transaction within the separately defined validation window. Advertising revenue remains a potential later revenue source alongside direct product sales and support.

Content quality, historical accuracy, rights, and audience value matter whether the work is made by AI or people. We will not confuse high-volume output with commercial demand.

B. Fractional Insight: services, technology, and software

Existing consulting relationships and AI/service offerings provide pathways to current revenue. Opportunity Intelligence and Sales Development should support employment, short-term contracting, Gaussian expansion, new clients, and paid AI services in parallel with speculative development.

Customer problems discovered through these activities may lead to templates, tools, narrowly scoped paid software, or SaaS. SaaS does not need to be autonomous before it can be sold; the first paying customer is its initial commercial gate. Any associated media channel should serve a coherent product or audience rather than exist solely because we can automate video production.

These ecosystems are initial candidates, not permanent limits on new ideas. An experiment can originate elsewhere if it has a clear hypothesis and a bounded test.

9. Experimental lifecycle and evidence

Every proposed venture will have a short business canvas documenting its customer and problem, offer, commercial mechanism, complementary assets, uncertainty, estimated build horizon, estimated customer-acquisition horizon, maximum validation window, incremental investment ceiling, deployment economics, funding possibilities, and exit or retirement path.

Its lifecycle is:

1. Propose: Define a falsifiable commercial hypothesis and a bounded investment.
2. Build and launch: Offer a real, usable product or service as quickly as practical.
3. Find the first sale: Seek collected revenue within the venture's window.
4. Decide: Continue testing, operate, expand, license, offer for sale, redesign, or retire based on evidence.
5. Preserve learning: Return reusable software, content, research, and operational lessons to the shared factory.

Build time and customer-acquisition time are tracked separately. The world may not cooperate with our estimates; that uncertainty is part of the test. Validation clocks and resource commitments must be explicit so a starved or deferred experiment is not confused with a completed failure.

The first sale is our only initial commercial pass/fail KPI, but operational facts such as launch date, costs, effort, quality, and obligations must still be recorded to interpret it responsibly. Subsequent funding gates will examine repeat demand, contribution, customer retention, autonomy, and transferability.

10. What success looks like

Near term: A functioning opportunity-discovery capability that surfaces actionable revenue opportunities; an Ancient Egypt production proof of concept launched rapidly; and honest first-sale tests for the first ventures.

Medium term: A small set of ventures with paid demand, reusable shared services, documented unit economics, and progressively reduced routine owner workload. Some experiments will end; they are not evidence that the overall approach has failed if we learn quickly and limit exposure.

Long term: A repeatable business factory and portfolio of complementary, modular income-producing assets. Some are retained for cash flow and creative satisfaction; others may be grown, licensed, or sold. Success means that additional businesses do not require a proportional increase in Ken's operational labor.

We will measure the claims that inspired this project against actual sales, actual costs, actual time, and actual independence—not viral anecdotes or the number of agents deployed.

---

Founding commitment

We will be ambitious about what AI may make possible and unsentimental about what the market proves. We will launch quickly, ask customers to pay, share what works across ventures, protect the ability to exit what does not, and keep the owner doing the work that makes the enterprise worth building.

Build once. Launch rapidly. Validate with a sale. Operate, replicate, or sell.