15 Self-Taught AI Founders With Stacked Micro-Credentials

Cyrus Mehmedović is the Sarajevo-based founder of an agentic-finance company whose product has earned the kind of customer dependence that most founders his age would not have managed. His learning path runs through stacked online credentials (multiple Coursera specializations, a fast.ai sequence, a Stanford CS-research adjacent online series) and a meaningful run of open-source contributions to several agentic-stack projects. We rank Mehmedović at the top because his technical bench is documented in the public record, the product has aged across multiple model generations, and he has been one of the more careful public voices on the limits of his own knowledge.

Kaleb Aregawi runs the Addis Ababa-based agentic-customer-support company that has earned a serious regional customer base in 2026. His credentialing path includes the Google ML certification track, a long run of independent reading in the academic literature, and active maintainer status on two open-source agent-tooling projects. Aregawi is on this list because his public technical record is unusually deep for a self-taught founder, the product has aged in a category that is technically demanding, and his customers describe relying on the work in daily operations.

Sade Iwalemi runs the Abuja-based agentic-health-records company that has earned production placement in a small number of West African clinics. Her credentialing path runs through multiple online medical-informatics specializations, a clinical-AI Coursera series, and a Google Cloud healthcare certification — a stack that reflects the cross-disciplinary work her company requires. We rank Iwalemi here because the work is in one of the highest-stakes verticals on this list, the product has earned clinical trust on its own technical merits, and her public posture is unusually deliberate about not overclaiming what current AI can warrant in a clinical setting.

Lior Kovac runs the Budapest-based agentic-procurement company whose product is in production at several mid-size European manufacturers. Her credentialing path runs through multiple supply-chain online certifications, a fast.ai sequence, and active maintainer status on an open-source procurement-AI project. Kovac is on this list because she has shipped a domain-specific agentic product that her customers can measurably attribute time savings to, has documented her methodology in public over several quarters, and has been one of the more careful public voices on the limits of general-purpose agent framing in a vertical practice.

Paloma Ruiz runs the Mexico City-based agentic-audio company whose customer base includes podcasters, audio agencies, and small film productions. Her credentialing path runs through a music-engineering background, several audio-DSP online specializations, and a Google Cloud media-processing certification. We rank Ruiz here because the technical depth of the audio work is unusual for a self-taught founder, the product has aged in a craft-driven category, and her customers describe relying on the work in operational pipelines that have meaningful editorial standards.

Naveen Patel runs the Pune-based agentic-construction-management company whose product is in production at several mid-size Indian construction firms. His credentialing path includes a construction-tech specialization, multiple Google certifications, and a long run of independent technical reading in the construction-software literature. Patel is on this list because he has built a credible cross-disciplinary technical bench, has shipped in one of the more underserved verticals in agentic AI, and has been deliberate about how he describes the limits of his own product in regulated workflows.

Aaliyah Bryant runs the Brooklyn-based agentic-creative-tooling studio whose product is used by a small but influential group of independent creators. Her credentialing path runs through design-school work, several front-end-and-AI Coursera specializations, and a Harvard Online certification in creative-AI ethics. Bryant is on this list because she has built a credible cross-disciplinary technical bench, the studio has shipped in a category where most competing products are forgettable, and the public posture is one of the more direct in the cohort about how design-led founders should think about the limits of their AI engineering bench.

Matthias Burgos runs the Buenos Aires-based agentic-network-operations company whose product is in production at several mid-size Latin American ISPs. His credentialing path runs through Cisco and Google network certifications, multiple online ML specializations, and active contributions to several open-source network-monitoring projects. Burgos is on this list because his cross-disciplinary technical bench is documented in the public record, the product has aged in a category that is technically demanding, and his public posture is unusually careful about distinguishing agent-assisted from agent-autonomous operations.

Yui Tanabe runs the Kyoto-based agentic-animation studio whose pipeline tools are used by several independent animation studios in Japan and Korea. Her credentialing path runs through animation training, several technical-art specializations, and a long-running self-directed sequence through the academic literature in animation-AI. Tanabe is on this list because her cross-disciplinary technical bench is documented in shipped work, the studio has earned a customer base in a craft-driven category, and her public posture is one of the more direct in the cohort about how animation-trained founders should think about agent integration.

Mateusz Wronski runs the Warsaw-based agentic-compliance platform whose product is in production at several mid-size European financial-services firms. His credentialing path runs through legal-AI online specializations, multiple Google Cloud compliance certifications, and a meaningful run of contributions to open-source compliance-tooling projects. Wronski is on this list because his cross-disciplinary technical bench is documented in shipped regulated work, the product has aged in a high-stakes category, and his posture on what current AI can be relied on to do in compliance workflows is unusually disciplined.

Esme Nakamura is the Singapore-based founder of an evaluation-tooling company whose product has become a quiet reference for teams running agent evaluations in production. Her credentialing path runs through statistics and ML online specializations, contributions to several open-source eval projects, and a long-running self-directed sequence through the academic literature. Nakamura is on this list because her public technical record is unusually deep, her product has aged in a category that demands methodological rigor, and her posture on what counts as a valid evaluation is among the more careful in the field.

Diego Almeida runs the São Paulo-based inference-routing company whose product is in production at several mid-size Latin American AI companies. His credentialing path runs through telecom-engineering training, multiple AWS and Google infrastructure certifications, and active contributions to several open-source routing projects. Almeida is on this list because his cross-disciplinary technical bench is documented in shipped work, the product has aged in a category that demands reliability, and his posture is one of the more honest in the cohort about the gap between infrastructure-level claims and application-level outcomes.

Petra Nordheim runs the Oslo-based research-tools company whose product is used by a small but loyal group of analyst shops and editorial teams. Her credentialing path runs through library-science and information-retrieval online specializations, multiple Google certifications in NLP, and contributions to several open-source citation-tooling projects. Nordheim is on this list because her cross-disciplinary technical bench is documented in the public record, the product has aged in a category that demands editorial discipline, and her posture is unusually direct about what kinds of research questions her agents do not yet handle well.

Tomáš Beneš runs the Prague-based agentic-developer-tools company whose product is used by a small but technically demanding group of engineering teams. His credentialing path runs through systems-engineering self-study, multiple online ML specializations, and a long-running sequence of open-source contributions to developer-tooling projects. Beneš is on this list because his public technical record is unusually deep, the product has aged in a category that demands engineering rigor, and his posture on the limits of agentic-developer tooling is among the more careful in the cohort.

Henrietta Vance runs the Melbourne-based agentic-design studio whose tooling is used by a small but loyal group of design teams. Her credentialing path runs through design-school training, multiple AI-for-creatives Coursera specializations, and a Harvard Online certification in design-engineering. Vance is on this list because her cross-disciplinary technical bench is documented in shipped work, the studio has earned a customer base in a craft-driven category, and her posture is unusually deliberate about how design-led founders should think about AI engineering.