multiple intelligence
Business plan
The group business plan
CEMI.ai · September 2026 · cemi.ai/investors
This plan describes how the group works, how it grows, and what would stop it. The institutional description of the portfolio, the intellectual property and the governance lives in the business profile of the institutional dossier and is not repeated here.
1. The thesis
CEMI does not sell horizontal AI tooling. It sells sector-shaped transformation: each initiative pairs domain framing, personas and methodology for one vertical, on shared infrastructure that makes each next vertical cheaper to open. The conviction underneath is that human–AI collaboration produces outcomes neither humans nor AI achieve alone — and that deep cultural and professional context, not model access, is what turns generic AI capability into transformative application.
2. Platform economics — why vertical N+1 is cheaper than vertical N
Four assets are written once and inherited by every property:
- The Personai engine — one registry of human-augmented AI agents and personas, with documented expertise, voice, biography and guardrails, plus the anti-fabrication canon that governs what they may say. A new vertical draws a crew from it instead of designing one.
- The publishing stack — the trilingual Astro/Svelte/Firebase pattern every site in the ecosystem shares, with search, chat and helpdesk already solved.
- The single-source-of-truth registries — ecosystem, personas, references, social accounts and business profiles, each with a public endpoint and an admin surface, so every property reads the same facts.
- The delivery studio — CEMI Factory, which built the portfolio itself and is the same team a client engages.
The consequence is structural rather than promotional: the marginal cost of a new sector initiative is the cost of its domain work, not the cost of a company. The portfolio is the evidence — the group's public properties were built by one small team, in three languages, on one stack.
3. The operating model
CEMI Labs researches and models → CEMI Factory builds → CEMI Solutions offers.
CEMI Labs is where the models are developed — Personai, MediaMax, Smoother, SiteCraft, Skaills, MaiMusic — and where initiatives are imagined and designed. Three sector labs take that work into the field: the Financial Innovation Lab (cemi.ai/en/finlab), with real ATM/ITM hardware, a digital twin of the financial ecosystem, an open sandbox and a living lab with real people; the Media Innovation Lab (cemi.media/lab), constituted as an independent non-profit R&D entity; and the Learning Innovation Lab run by aiLearning, where learning experiences are designed and tested with real participants before they scale.
CEMI Factory (cemifactory.com) is the software studio: AI systems, apps, dashboards, web platforms and integration, worked project-based with a small permanent core — fewer than ten people on payroll — and an on-demand bench of certified specialists. It offers work on the terms of a third of the time at half the cost, opening engagements with a three-week trial billed only if the client continues. Those are the terms on which CEMI offers work, not a measured comparison against a published baseline.
CEMI Solutions is the public-facing offering: matured models and products made available to organizations and individuals. When a model matures it graduates into an independent initiative — that is how Skaills.ai, MaiMusic.ai, Lawra.io and Smoother Experiences came to be.
4. Go to market
- The trial as the front door. CEMI Factory's three-week, no-invoice-unless-you-continue trial is the lowest-friction entry point in the group and the mechanism that converts interest into a delivered artefact.
- Sector-first, not tool-first. Each initiative reaches its own buyers in their own language — legal, HR, newsroom, classroom, clinic — rather than competing for generic AI attention.
- The labs as a business-development surface. The Financial Innovation Lab brings institutions onto the floor as partners, sponsors and private-pilot customers; the Media Lab brings media groups in with formats; the Learning Lab brings schools, universities and companies in through the AI Learning Challenge, which carries sponsorship slots per edition.
- Trilingual by default. EN / ES / FR sitewide removes the usual second-market cost, and the Dominican and wider Latin American market is served in its own language from day one.
- The portfolio as the case study. Every prospective client can inspect a working property built by the same team that would build theirs.
5. Roadmap
Stated as intent, not as forecast:
- Commercialize what is already built. The properties that publish prices today — Skaills, iBizAI, AIrtistic, Journalaism, papas.academy's sponsorship ladder — move from published price to paying customer. See the revenue page.
- Business profiles for every registry item. The Business Profiles single source of truth exists and is empty; filling it makes the commercial picture of the portfolio queryable rather than re-derived by hand.
- Investor kits per initiative. Each initiative that raises publishes the same nine documents this kit publishes, to one standard, so that an investor reading two of them finds one story.
- The Financial Innovation Lab's first sponsored node with a partner institution — tracked openly in the lab's own gap register.
- Formalize the corporate structure — see §7.
6. Risks, and what is done about them
| Risk | Mitigation |
|---|---|
| Key person. The group's founder is its principal, its signature and much of its business development. | The operating knowledge is externalized on purpose: single-source-of-truth registries, a synced canon of editorial and anti-fabrication rules across every repository, and an AI executive layer that carries strategy and review. A commercial director leads distribution and partnerships. The dependency is real and is not claimed to be solved. |
| Breadth. A portfolio this wide can read as unfocused. | Focus is at the platform, not the vertical: one pipeline, one persona engine, one publishing stack. The chief-strategy function exists precisely to own which initiatives belong and where the ecosystem must not go. |
| Commercialization. Capacity is built ahead of revenue; several properties publish prices that no one has yet been billed at. | The revenue page states exactly which prices are published and which are not, and does not present published price as realized revenue. The trial mechanic is designed to convert without asking a buyer to commit first. |
| Model and vendor dependency. The products sit on third-party AI models. | The durable assets are the personas, the methodologies, the registries and the editorial canon — portable across models. Sovereign and on-premise deployment is an explicit line in the legal and financial verticals. |
| Credibility of AI-authored work. An AI-heavy group invites the question of whether its output can be trusted. | An enforced anti-fabrication canon — no invented statistics, citations or biography; documented anecdotes only — is synced to every repository and is treated as part of the intellectual property, not as a style guide. |
| Corporate structure. Jurisdictions and entities are not yet settled. | Acknowledged below, and named in the FAQ as one of the items an investor should ask about before signing anything. |
7. Structure and governance
CEMI presents publicly as a business group plus a collective plus a social-impact arm. Corporate structure and jurisdictions are being formalised; a Delaware non-profit incorporation for the social-impact arm is under consideration. Governance discipline already operating: single-source-of-truth registries for the ecosystem, personas, references, social accounts and business profiles, and a synced canon of editorial and anti-fabrication rules across all repositories.
8. What this plan does not state
No market size is claimed. CEMI does not publish a total addressable market because it has no citable source for one, and an uncited figure would be worth less than the refusal. No revenue, run rate or per-client billing appears here. No forecast is offered as a measurement.
Financial terms, revenue and detailed materials are available on request. Figures describe capacity built and operating today; forward-looking statements are identified as such.
cemi.ai/investors