Voice AI Innovators: Disrupting Overlooked Markets Outside of Goldman and Meta

Revolutionizing Voice AI in Africa and the Middle East

Customer support and service are currently at the forefront of the voice AI industry, with a particular focus on Africa and the Middle East. However, creating a product that emulates human interaction and responds seamlessly in these regions presents unique challenges. Many established players in the field have overlooked the specific requirements of these markets.

AethexAI, a promising startup launched last year, has secured $3 million in pre-seed funding led by 4DX Ventures, with contributions from Enza Capital, Dorm Room Fund, Mojo Ventures, and Stanford GSB 26 Fund. Notable individual investors include Stanford faculty members, telecom executives, and AI experts from Anthropic.

Unlike other companies that rely on existing orchestration tools like Vapi and LiveKit, AethexAI has developed its own customized model and orchestration layer from scratch. This unique approach enables the platform to effectively cater to the diverse dialects of English, French, and Arabic spoken across the target markets.

The startup is now introducing its platform to enterprises for testing and adoption, offering APIs and SDKs for developers to explore its models.

Founded by Mariama Diallo and Ayooluwa Odemuyiwa, AethexAI aims to address the specific needs of emerging markets. Diallo, with experience at Goldman Sachs and YC-backed ModelML, serves as CEO, while Odemuyiwa, a Caltech graduate and former Meta employee, acts as CTO. Their shared vision for creating solutions for underserved markets led them to establish the company.

While businesses worldwide are increasingly adopting AI tools for automation, success is not guaranteed. In Egypt, for example, a call center attempted to automate a significant portion of its calls but ultimately reverted to human intervention due to unsatisfactory outcomes. Similarly, support centers across Africa struggle to find skilled engineers for cost-effective call automation, highlighting a persistent challenge in the region.

Odemuyiwa emphasized the importance of minimizing latency and ensuring efficient call automation. By developing small, locally hosted models through the Kora series, ranging from 300 million to 1.7 billion parameters, AethexAI effectively addresses the latency issue without compromising accuracy.

To train these models, the startup leveraged anonymized recordings from a call center partner and distributed hard drives to radio stations in Africa for additional audio data collection. Utilizing a contributor network of university students for data annotation and pronunciation validation, AethexAI now handles over 17,000 calls daily.

On the operational front, the company prioritizes guiding new voice AI clients through the implementation process, offering on-site demonstrations and workshops to identify optimal automation use cases.

While AethexAI is open to collaborations across various industries, its current focus involves applications such as debt collection, customer activation, and KYC verification. The company recruits forward-deployed engineers on contractual terms to serve local markets and establishes channel partnerships with telecom providers for voice AI call telephony solutions.

Walter Baddoo, co-founder of 4DX Ventures, underscores the distinct nature of the Africa and Middle East market compared to traditional voice AI markets. Enterprises in these regions experience significantly higher call volumes and require systems tailored to local dialects, code-switching, and informal speech patterns, as well as compatibility with existing telephony infrastructure.

While global voice AI companies expand rapidly, startups like AethexAI recognize the need for specialized models, local partnerships, and region-specific infrastructure. These initiatives address market gaps that major players are unable or unwilling to bridge.