How to Actually Get Hired

Mariama Diallo's path to founding a voice AI company in London started in investment banking at Goldman Sachs, where she was an Associate in the Cross Markets Group. The Cross Markets Group works across asset classes and geographies, which means she was thinking in systemic terms about how capital flows across markets rather than working on a single sector. She then went to Wharton for her MBA (Class of 2023), where she was a regular attendee of the Wharton Africa Business Forum before COVID, and wrote publicly about choosing Wharton specifically because of its commitment to the African continent as a viable business hub: "Wharton had a diverse student body and allowed them to promote not only the African continent as a viable business hub, but also to share the continent's diverse cultures and values with other students on campus." That framing, Africa as a legitimate hub rather than a development project, is the intellectual foundation of AethexAI's commercial positioning. After Wharton, she joined ModelML as its first product and growth hire. ModelML is a YC-backed machine learning platform company that helps enterprises build production AI products. Being the first product and growth hire at a YC company means she was doing customer discovery, pricing, GTM design, and product feedback loops simultaneously, for a product with enterprise clients, before she had a large team. That is the same role she plays at AethexAI: guiding new clients personally through automation decisions via in-person demonstrations and workshops before scaling any deployment. Her and Ayooluwa's fieldwork pattern is what makes the founding story credible. They spent time on the ground with businesses across Africa and the Middle East, where they repeatedly encountered the same problem: voice AI products that worked in demos but collapsed in production. They left their respective roles in San Francisco and relocated to London to build a purpose-designed voice infrastructure platform from scratch. That geographic decision, London over SF, is deliberate: closer timezone overlap with Africa and the Middle East, access to the diaspora talent pool, and lower build costs than the Bay Area. Her quote from the TFN coverage is the product thesis in its most compressed form: "Voice is already how businesses operate across emerging markets, but the technology behind it hasn't kept up. We kept hearing the same thing from customers: that existing tools simply didn't work in their environments. That's why we built our own model stack and infrastructure from the ground up, designed for how these markets actually operate."

Mariama Diallo's path to founding a voice AI company in London started in investment banking at Goldman Sachs, where she was an Associate in the Cross Markets Group. The Cross Markets Group works across asset classes and geographies, which means she was thinking in systemic terms about how capital flows across markets rather than working on a single sector. She then went to Wharton for her MBA (Class of 2023), where she was a regular attendee of the Wharton Africa Business Forum before COVID, and wrote publicly about choosing Wharton specifically because of its commitment to the African continent as a viable business hub: "Wharton had a diverse student body and allowed them to promote not only the African continent as a viable business hub, but also to share the continent's diverse cultures and values with other students on campus." That framing, Africa as a legitimate hub rather than a development project, is the intellectual foundation of AethexAI's commercial positioning. After Wharton, she joined ModelML as its first product and growth hire. ModelML is a YC-backed machine learning platform company that helps enterprises build production AI products. Being the first product and growth hire at a YC company means she was doing customer discovery, pricing, GTM design, and product feedback loops simultaneously, for a product with enterprise clients, before she had a large team. That is the same role she plays at AethexAI: guiding new clients personally through automation decisions via in-person demonstrations and workshops before scaling any deployment. Her and Ayooluwa's fieldwork pattern is what makes the founding story credible. They spent time on the ground with businesses across Africa and the Middle East, where they repeatedly encountered the same problem: voice AI products that worked in demos but collapsed in production. They left their respective roles in San Francisco and relocated to London to build a purpose-designed voice infrastructure platform from scratch. That geographic decision, London over SF, is deliberate: closer timezone overlap with Africa and the Middle East, access to the diaspora talent pool, and lower build costs than the Bay Area. Her quote from the TFN coverage is the product thesis in its most compressed form: "Voice is already how businesses operate across emerging markets, but the technology behind it hasn't kept up. We kept hearing the same thing from customers: that existing tools simply didn't work in their environments. That's why we built our own model stack and infrastructure from the ground up, designed for how these markets actually operate."

Ayooluwa Odemuyiwa holds a BS in Computer Science from the California Institute of Technology (Caltech), focused on Machine Learning and Computer Vision. Caltech is one of the most technically rigorous undergraduate institutions in the world, with an undergraduate student body smaller than most graduate programmes. She graduated into a software engineering role at AeroVironment, a defence technology company that builds drones and unmanned aerial systems for military applications, where she built mission-critical ground control software. Ground control software for drones is not forgiving: the systems must work reliably under communication degradation, packet loss, and real-world network conditions — precisely the infrastructure constraints that Kora is designed to handle in African telecom environments. She then joined Meta as a Software Engineer, building systems used by billions. She was also a Teaching Assistant for graduate-level machine learning coursework at Caltech, and an AI/ML Fellow at TechnoServe, the international NGO, focused specifically on AI applications for coffee farmers in Central America, which is an early signal of her interest in deploying AI in contexts that are not Silicon Valley mainstream. She enrolled at Stanford Graduate School of Business, where she connected with the Stanford community that later made the 26 Fund investment in AethexAI, before leaving to co-found the company. Her quote in the TFN coverage is the most technically precise public description of why the existing voice AI market fails in these environments: "Voice AI failed in these markets at every layer of the stack. Latency, cost, poor handling of code switching, and weak performance under packet loss, jitter, and low-bitrate audio in real telecom networks led these systems to break in production. The fix was not incremental. It required redesigning the entire stack. Kora 1 is our family of speech models, specialised by dialect and fully self-hosted." The AeroVironment credential deserves specific attention. Building software that controls drones in the field, where the cost of a failure is physical and immediate, requires rigorous testing against degraded communication scenarios and fault-tolerant architecture. Those engineering instincts are not common in voice AI infrastructure companies. They are precisely what allows AethexAI's stack to perform at <500ms latency under the packet loss and jitter conditions that defeated every prior attempt in these markets. Lesser-known facts: Her TechnoServe fellowship, working on AI applications for coffee farmers in Central America, is the detail most people skip. That role required building AI systems that are useful and deployable in environments without reliable connectivity, with users who interact in Spanish and indigenous languages, for agricultural decisions with real economic stakes. The skill set for that is almost identical to the skill set for deploying voice agents in Francophone Africa or Arabic-speaking Egypt. She was solving the emerging market AI deployment problem in Latin America before she applied it to voice AI in Africa and the Middle East.

Ayooluwa Odemuyiwa holds a BS in Computer Science from the California Institute of Technology (Caltech), focused on Machine Learning and Computer Vision. Caltech is one of the most technically rigorous undergraduate institutions in the world, with an undergraduate student body smaller than most graduate programmes. She graduated into a software engineering role at AeroVironment, a defence technology company that builds drones and unmanned aerial systems for military applications, where she built mission-critical ground control software. Ground control software for drones is not forgiving: the systems must work reliably under communication degradation, packet loss, and real-world network conditions — precisely the infrastructure constraints that Kora is designed to handle in African telecom environments. She then joined Meta as a Software Engineer, building systems used by billions. She was also a Teaching Assistant for graduate-level machine learning coursework at Caltech, and an AI/ML Fellow at TechnoServe, the international NGO, focused specifically on AI applications for coffee farmers in Central America, which is an early signal of her interest in deploying AI in contexts that are not Silicon Valley mainstream. She enrolled at Stanford Graduate School of Business, where she connected with the Stanford community that later made the 26 Fund investment in AethexAI, before leaving to co-found the company. Her quote in the TFN coverage is the most technically precise public description of why the existing voice AI market fails in these environments: "Voice AI failed in these markets at every layer of the stack. Latency, cost, poor handling of code switching, and weak performance under packet loss, jitter, and low-bitrate audio in real telecom networks led these systems to break in production. The fix was not incremental. It required redesigning the entire stack. Kora 1 is our family of speech models, specialised by dialect and fully self-hosted." The AeroVironment credential deserves specific attention. Building software that controls drones in the field, where the cost of a failure is physical and immediate, requires rigorous testing against degraded communication scenarios and fault-tolerant architecture. Those engineering instincts are not common in voice AI infrastructure companies. They are precisely what allows AethexAI's stack to perform at <500ms latency under the packet loss and jitter conditions that defeated every prior attempt in these markets. Lesser-known facts: Her TechnoServe fellowship, working on AI applications for coffee farmers in Central America, is the detail most people skip. That role required building AI systems that are useful and deployable in environments without reliable connectivity, with users who interact in Spanish and indigenous languages, for agricultural decisions with real economic stakes. The skill set for that is almost identical to the skill set for deploying voice agents in Francophone Africa or Arabic-speaking Egypt. She was solving the emerging market AI deployment problem in Latin America before she applied it to voice AI in Africa and the Middle East.