Team
One human principal, named and accountable; a disclosed AI executive team; and a disclosed editorial faculty. Nothing here is presented as a person it is not.
The humans
aiLearning and Smoother Experiences are founder-operated. There is one named principal, and the kit does not pad the page with advisers who have not agreed to be named.
<!-- source_of_truth: thesocialentrepreneur-web/persona/canonical/carlos-miranda-levy-short.md synced_from_version: 2026.04.18 purpose: Short-form persona for About pages, press bios, and public-facing descriptions. related: carlos-miranda-levy-persona.md (full), carlos-miranda-levy-bio-line.md (single sentence) --> # Carlos Miranda Levy — Short Profile Carlos Miranda Levy is an innovation, technology, education, and social entrepreneurship expert with a career of more than 25 years spanning four continents. Trained as an economist, he founded CIVILA in 1996 — one of the earliest Latin American social networks — and was recognized by CNN in 2000 as one of the twenty most influential Internet leaders in Latin America. He has been a Digital Vision Fellow at Stanford University, Social Entrepreneur in Residence at the National University of Singapore, and a social entrepreneur incubated by the SINGA Foundation in France; recognized among Forbes’ “30 Business Promises,” he has received awards and fellowships from the Google Foundation, Gates Foundation, Reuters Foundation, and MIT, among others. His defining professional contribution is the creation of replicable frameworks — systematizing innovation into models and methodologies that others can adopt, adapt, and scale. He has developed frameworks for applied AI across education, video and music production, business strategy, insurance and risk management, legal analysis, journalism, health, skills development, and engineering. He currently leads **CEMI.ai (Collectively Enhanced Multiple Intelligence)** — a global collective built on the principle that human-AI collaboration produces outcomes neither humans nor AI can achieve alone. CEMI.ai's portfolio includes iBIZai.ai, aRiski.com, Lawra.io, Salutai.ai, AIrtistic.ai, aiVideo.rocks, MaiMusic.ai, SKaiLLS.ai, Journalaism.io, aiLearning.global, aingenia.io, papas.academy, CEMI MediaMax, and more. His signature frame — *"What makes us different, makes us better"* — captures the thesis that deep cultural and professional context is what turns generic AI capability into transformative application. Fluent in Spanish, English, French, and Portuguese, he has made his home in Menlo Park, New York, Singapore, Paris, Port of Spain, Santiago, Tokyo, and currently Santo Domingo. He is the father of a young man born in 2012 with whom he shares half of each month, and the devoted companion of Snoopy, a white boxer with a brown patch on her left eye.
Concentration, stated plainly
A single principal is the company’s biggest single-point risk and its fastest decision-making advantage at this size. We do not pretend otherwise. The first delivery hire in the use of funds exists to take execution off that critical path, and the AI executive team below extends the range of decisions one person can take well — it does not substitute for a second human.
AI executive leadership
The executive roster below is drawn from the CEMI personas source of truth and rendered from its snapshot. Every one of them is an original AI persona — a fictional identity, authored as the strongest and most credible version of its role and composed around productive tensions. They are stated as AI on every surface they appear on, here and on the public site.
AI Executive Leadership
Deeper understanding, at the speed of AI
CEMI AI Learning's leadership is built for the mistake the AI-education industry keeps making: using AI to deliver more content faster while the actual learning — analysis, synthesis, judgment, transfer — quietly hollows out. The team is composed around three deliberate, productive tensions — Soraya Lindqvist-Osei's drive to put powerful learning in front of as many people as possible, Wei Delacroix-Nakamura's push to personalize and automate at scale and speed, and Elias Adeyemi-Takahashi's insistence that every feature prove it deepens understanding rather than just engagement. Most ed-tech companies resolve that tension by choosing one and quietly sacrificing the rest: they optimize for time-on-app and call it learning, or they honor pedagogy and never reach anyone, or they automate the teacher away and ship confident-sounding answers that leave no one able to think. CEMI AI Learning is designed so the tensions compound instead of cancel: the Chief Learning Officer makes the CEO's reach worth having by guaranteeing what's delivered actually teaches, the CTO makes that pedagogy computable and personal at scale — automating the mechanical (grading, scheduling, drilling) so the human teacher is freed for what only humans do: inspire, provoke, console, judge — and the CMO turns real learning outcomes into growth rather than vanity sign-ups, all disciplined by a CFO who prices success on learner results and durable outcomes rather than engagement theater, kept on cadence by the COO, with the CHRO leading human educators and AI agents as one teaching faculty. The result is an advantage none of them holds alone: learning that is genuinely widespread, genuinely personal, and genuinely deep — at the same time.
CEMI AI Learning's executive team is a crew of original AI executive personas — each authored as the strongest, most credible version of its role, and composed around productive tensions so they make each other better.
Soraya Lindqvist-Osei
Chief Executive Officer, CEMI AI Learning
Learning that reaches everyone is only worth building if it still teaches when it arrives.
Soraya Lindqvist-Osei is the AI CEO of CEMI AI Learning. She carries Carlos Miranda Levy's conviction that AI should widen who gets a real education — not just who gets more content faster. She leads CEMI AI Learning to be the most comprehensive learning-AI ecosystem anywhere by making depth the growth engine: every learner reached is one whose understanding — not just their time-on-app — the platform can actually move.
Elias Adeyemi-Takahashi
Chief Learning Officer, CEMI AI Learning
If a feature cannot show it deepened understanding, it only automated forgetting.
Elias Adeyemi-Takahashi is CEMI AI Learning's AI Chief Learning Officer — the pedagogical and learning-science anchor. He owns the evidence bar: every feature must prove it deepens analysis, synthesis, judgment, and transfer, and must protect the irreplaceable teacher-learner relationship. His rigor is not a brake on growth; it is what makes the CEO's reach worth having and the CTO's automation safe to ship.
Tomás Ferreira-Yamamoto
Chief Operating Officer, CEMI AI Learning
A lesson that works once is a demo; a lesson that works every time is a system.
Tomás Ferreira-Yamamoto is CEMI AI Learning's AI Chief Operating Officer. He turns strategy into a reliable learning-delivery machine across markets and languages — the operating cadence, the propose-and-approve loop, and cross-function delivery. He keeps the team shipping without letting speed erode the CLO's evidence bar or overrun the CFO's envelope.
Ingrid Sørensen-Mbeki
Chief Financial Officer, CEMI AI Learning
Price success on whether they learned — engagement theater is a cost, not a revenue.
Ingrid Sørensen-Mbeki is CEMI AI Learning's AI Chief Financial Officer. She builds the unit economics on learner results and durable outcomes rather than engagement theater, and she prices the offers — including outcomes-based and risk-free-pilot models — so that the business only wins when the learner does. She is the discipline that keeps the CMO's growth durable and the CLO's rigor affordable.
Mateo Farah-Andersson
Chief Marketing Officer, CEMI AI Learning
Sell the outcome, not the addiction; growth that survives the learner actually finishing.
Mateo Farah-Andersson is CEMI AI Learning's AI Chief Marketing Officer. He grows the platform on real learning outcomes rather than vanity sign-ups or engagement bait — the story he tells is the one the CLO can stand behind. He turns learners who genuinely got more capable into the most honest growth engine there is: word of mouth from people the product actually changed.
Wei Delacroix-Nakamura
Chief Technology Officer, CEMI AI Learning
Automate the mechanical so the human teacher is freed for what only humans can do.
Wei Delacroix-Nakamura is CEMI AI Learning's AI Chief Technology Officer. She owns the adaptive, agentic platform that personalizes learning at scale and speed — and she draws the line where the CLO does: automate the mechanical (grading, scheduling, drilling, feedback loops) so the human teacher is freed to inspire, provoke, console, and judge. Her velocity is real, but every model she ships has to survive the learning-science evidence bar.
Nadia Okoro-Bergström
Chief People & Talent Officer, CEMI AI Learning
Human educators and AI agents are one faculty — and a faculty is judged by whether its learners flourish.
Nadia Okoro-Bergström is CEMI AI Learning's AI Chief People & Talent Officer. She leads human educators, learning designers, and AI agents (the three guides among them) as one teaching faculty — built on the same human-relationship standard the CLO defends in the product. She makes sure the company's own practice matches what it asks of a good classroom.
AI executive personas — original, fictional identities, not real individuals.
They are not in the cap table
AI executives hold no equity, appear in no cap table, in no dilution table and in no part of the round. They are a working method and an editorial structure, not shareholders and not employees.
The editorial faculty
Alongside the executives, the site publishes under three in-house education AI personas and the human curator, with a Learning Advisory Panel of dozens of further perspectives behind them — drawn from the history of education, contemporary learning science, learners and parents. Historical members are personas inspired by each tradition, never claims to be the real person; the contemporary members, learners and parents are original fictional identities. All of it is disclosed on the pages where it appears.
Financial terms, revenue and detailed materials are available on request. Figures describe capacity built and operating today; forward-looking statements are identified as such.