Business plan

The problem we are in

Institutions are not short of AI tools. They are short of a defensible answer to three questions: what our people actually need to be able to do, how we get them there without pretending the technology is settled, and how we show anyone — a board, a ministry, a parent — that it worked.

The industry’s reflex is to answer with more content, delivered faster. That is the mistake this company is built against. Using AI to push more material through the same pipe makes the delivery cheaper and quietly hollows out the part that mattered — analysis, synthesis, judgement, transfer. Our editorial position is the opposite one, and it is not a slogan: the learner’s experience is the subject, teaching is in service of it, and no feature earns its place until it can show it deepened understanding rather than engagement.

The practical consequence is the shape of the business. A consultancy that leaves behind a report is selling the wrong artefact. A platform sold to an institution that has not been taught what to do with it is shelfware. What an institution can actually use is the pair.

Two product lines

Product line 1 — aiLearning: the open site, the programmes and six services

The open side is large and free on purpose: guides and learning paths, a prompt library, a glossary and tool directory, ethics scenarios, a readiness self-assessment, browser learning games, AI tutors, an AI-and-schools section written for Dominican educators and families, and long-form Insights published in five series. None of it is gated, and it is the reason an institution will take a first meeting.

The commercial side is six productized services, each with its own published page, scope and scoping process: AI Suite for Education, AI Readiness Assessment, Consulting, Personalized Training, Digital Twins for Education and Speaking Engagements. Two publish an investment band; the other four publish a scoping statement rather than a price, and the revenue page quotes each of them word for word.

The standing editorial voices — three in-house education AI personas plus the human curator, and behind them a Learning Advisory Panel of dozens of perspectives drawn from the history of education, contemporary learning science, learners and parents — are disclosed as AI throughout and are what lets the site publish at volume without publishing filler.

Product line 2 — Smoother Experiences: the learning platform

Smoother Experiences (experiences.ailearning.global) is the learning management platform the programmes run on. It is multi-tenant: an institution gets an academy under its own domain, palette, typography and logos, with a curated catalogue, its own managers and roles, enrolment and approval workflows, analytics and certificates it controls.

It carries ten content types, a diagnostic layer, learning paths and series that group programmes into certified pathways, a transactional email system, a public catalogue API any partner site can read the catalogue through, and an AI Learning Designer that authors programme content with the same engine the platform runs on — with a free public tier.

It is WCAG 2.1 AA compliant, has role-based access rules, lets a user export all their data, and gates a programme on a published quality bar before it ships. Those are not consumer features. They are the reasons a training director or a ministry can say yes.

What it does not do: it issues certificates on completion, with a verification code and a public verification page, but it does not sell a certification and claims no accreditation anywhere. This plan makes no certification revenue claim.

Why the two compound

The services carry the cost of getting inside an institution and are paid for doing it. The platform is what remains when the engagement ends, and it renews without consuming delivery time. Each service sold is a reason to deploy an academy; each academy deployed is a reason the next service is easier to sell into the same institution.

The programme library is the asset both lines draw on. A programme authored once is delivered by the training line, hosted on the platform, curated into a white-label catalogue, and read through the catalogue API by a partner site — four surfaces, one authoring cost.

The open material sits under all of it. It is what makes an unsolicited approach welcome instead of ignored, and it is the single largest reason the pipeline exists at all.

The model

Revenue takes four shapes: services sold per engagement; training and events sold per cohort or per booking; platform subscriptions and licences for a white-label academy; and sponsorship and grants for programmes that are free to the participant by design. The revenue page sets out all ten streams, each with its buyer, its income source and its published price or the words «on request».

The strategic move this round funds is the shift of weight from the first shape to the third. Services pay today and do not compound; a platform renewal does. Nothing here proposes to stop selling services — they are how the platform gets in the door — but the plan is judged on whether the recurring line grows.

Sponsored and public-sector programmes stay free to the participant permanently. That is stated on the platform’s own pages and is not a concession made to investors; it is a condition of the work.

Every revenue stream, its buyer and its published price are set out in full on the revenue-streams page. Revenue streams →

Go to market

The institutional route runs: open material → readiness self-assessment → paid assessment or consulting engagement → suite or training deployment → white-label academy. Each step is published and each is a real product, so a buyer can stop at any of them and still have bought something whole.

Speaking is the top of that route and is priced as reach rather than revenue, with reduced rates published for public-sector, academic and mission-aligned bookings.

The AI Learning Challenge — a live competition of creativity and problem-solving with AI, for learners from primary school to university, first edition scheduled for November 2026 — is the sponsor-funded route into school systems. There is no participation fee and none is planned: every tool a team uses must be free and available to everyone.

Three languages are not a marketing feature. English, Spanish and French are the operating languages of the institutions this is sold into, and the programme library, the platform and the site all carry all three.

On market sizing

This plan states no total addressable market. Figures for “AI in education” circulate widely and none of them is citable for the segment these two properties actually sell into — institutional AI capability-building, in three languages, bought by an education or learning function. Rather than publish a number we cannot source, we publish none. Where a citable figure exists, it will appear here with its source next to it.

The plan this round is sized to

US$250,000 over 18 months: 48% into the platform and the infrastructure under it, 52% into the programmes, the delivery capacity and the institutional reach. The line items and their monthly amounts are on the ask page.

Forward-looking. It is a plan for the money, identified as such, not a projection dressed as a record.

Risks we would rather state than have found

Concentration on one principal. The company is founder-operated. The AI executive team and the editorial personas extend what one person can publish and decide with; they do not replace the dependency, and the delivery hire in the use of funds is the first structural answer to it.

The platform is not yet priced. A pricing model exists as an internal draft and is explicitly unsettled. Until it is decided, the white-label line is sold on request, and this kit refuses to publish a band from a draft.

Sponsored work is lumpy. Public-sector and sponsor-funded programmes are the highest-credibility line and the least predictable one. They are treated as reach and evidence, never as the base case.

Model dependence. Both properties sit on top of frontier models we do not own. The methodology, the programme library and the editorial position are ours; the inference is rented, and the plan carries its cost as a line rather than assuming it away.