Editor’s note: African-Startups is a sister publication of EU-Startups, bringing trusted coverage of startups, venture capital, and innovation across Africa.
AethexAI, a startup building end-to-end voice AI infrastructure for emerging markets, has come out of stealth with a $3 million pre-seed round led by 4DX Ventures. The round also saw participation from Enza Capital, Dorm Room Fund, Mojo Ventures, and Stanford GSB 26 Fund, with angel participation from Stanford faculty, telecom executives, and AI researchers from Anthropic.
The AI startup was co-founded by Mariama Diallo, who serves as CEO, and Ayooluwa Odemuyiwa, who serves as CTO. Diallo brings a background spanning investment banking at Goldman Sachs and product and growth roles at YC-backed ModelML, while Odemuyiwa, a Caltech graduate, came to the venture from Meta and a stint at Stanford Business School.
According to TechCrunch, the two set out to address a gap they observed first-hand: in Egypt, a call centre that had automated a significant share of its calls rolled the system back because of poor results, while several support centres in Africa told them that finding and hiring engineers to automate calls at the right cost was a persistent headache.
The company’s core product is Kora 1, a proprietary model stack trained on real conversational speech with human-labelled data across emerging markets. Kora 1 comprises three components: Kora Read, an automatic speech recognition system built with proprietary datasets and human-assisted labeling for multilingual, code-switching, and noisy, low-connectivity environments; Kora Speak, a text-to-speech engine offering over 100 natural voices with dialect and accent-level localisation; and a real-time voice orchestration layer with turn-taking, interruption handling, and end-of-speech detection. A no-code enterprise platform and developer APIs and SDKs round out the offering.
Rather than using existing orchestration tools, the company built its own small models and orchestration layer from scratch to handle the localised dialects of English, French, and Arabic spoken across its target markets.
“The latency and jitter that we saw on automated calls in this region were outrageous. If we had become orchestrators, we might have had to use large models that were hosted outside the region, resulting in higher latency. We realised that in order for this to work, we have to use very small models and cut latency at every step,” said Odemuyiwa.
To train its Kora models, which range from 300 million to 1.7 billion parameters, the startup used anonymised recordings from a call centre partner and shipped hard drives to radio stations across Africa to collect more audio data, while building a contributor network of university students to annotate data and pronounce local names.
The company is already live in production, with one deployment processing over 15,000 outbound calls daily. According to 4DX Ventures, the platform is now handling more than 17,000 calls per day across support, sales, onboarding, collections, and other customer workflows.
“We always tell customers that we cannot be everything for everybody right now. We’re small. When we start talking to a company, we ask them to pick one use case that is the most important to them to start [with],” said Diallo. She highlighted current use cases such as debt collection, customer activation, and KYC (Know Your Customer) verification for banks and telecoms.
The investment is a bet on a market that has largely been underserved by other major AI firms. Walter Badoo, co-founder and managing partner of 4DX Ventures, says enterprises across Africa and the Middle East handle roughly three times the call volume of their Western counterparts and voice remains the dominant channel for customer interaction.
He further adds that existing systems were built for Western markets characterised by high-end GPU infrastructure, standard English and European speech environments, and enterprise workflows common in the United States and Europe.
“That creates real gaps when enterprises need systems that handle dialects, code-switching, and informal speech patterns, and that work within their existing telephony infrastructure and their actual price points,” Badoo said.
The company is also building channel partnerships with telecom providers to handle telephony for voice AI calls, hiring forward-deployed engineers on a contract basis to serve local markets. The startup sees localisation as its moat and that the infrastructure gaps the incumbents have left behind are too deep for a global rollout to fill.
