AethexAI, a UK-based company building voice AI infrastructure for emerging markets, has raised $3 million in a pre-seed round led by 4DX Ventures.
The startup also launched its platform today.
The round included participation from Enza Capital, Dorm Room Fund, Mojo Ventures, 26 Fund, and strategic angel investors — including Stanford faculty, telecom executives, and AI researchers from Anthropic.
Why voice AI keeps failing in Africa and the Middle East
Voice remains a primary channel for enterprise customer interaction across emerging markets. But many companies that experimented with voice AI found that existing tools didn’t work reliably in production.
The problems are familiar to anyone who’s tried to run a cloud-based speech system on a shaky mobile network: unreliable connectivity, fragmented telephony, high pricing, and poor handling of local speech patterns. In many cases, the solutions ended up costing more than human agents.
AethexAI says it rebuilt the voice stack from the ground up for these environments. The platform is self-hosted, uses market-localised models trained on proprietary data, and is wired directly into fully managed telephony, orchestration, and existing workflows.
It’s delivered through a no-code interface and APIs.
Related: CommerceClarity buys AI firm Katalogo
The company claims businesses can deploy and scale voice agents at a fraction of the cost of existing providers.
Targeting 1.5 billion people — but starting small
The startup is initially targeting a market of 1.5 billion people across Africa and the Middle East, where global providers have yet to deliver at scale. Plans include expanding to other emerging markets later. It was founded by Mariama Diallo and Ayooluwa Odemuyiwa. Diallo previously worked in investment banking at Goldman Sachs before joining YC-backed Model ML as its first product and growth hire. Odemuyiwa trained as a computer scientist at Caltech, built systems across aerospace and at Meta, then attended Stanford Graduate School of Business.
The pair started building last year after spending time on the ground with businesses across Africa and the Middle East. They saw that existing voice AI models weren’t built for these environments. Companies couldn’t automate large parts of customer interactions, leaving efficiency gains and revenue untapped.
They left their jobs to build the startup full-time.
Kora 1: the proprietary voice model stack
The platform runs on Kora 1, AethexAI’s own voice model stack.
It’s trained on licensed datasets from call centres, radio, and content platforms. The firm says it’s designed for noisy environments, multiple accents, and languages. A developer platform is also launching, letting third parties build voice applications across the region using a single API. That’s a move that could help spread adoption, though it also means the startup will need to support a wide range of use cases — something that’s tripped up bigger players. Diallo said the company kept hearing the same thing from customers: 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.”
Odemuyiwa noted that voice AI failed at every layer of the stack in these markets. “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.” She argued the fix wasn’t incremental — it required redesigning the entire stack.
Related: DJI OM 5 The Most Well-refined Smartphone Gimbal Stabilizer
We built and own the data pipeline behind them.
“Kora 1 is our family of speech models, specialised by dialect and fully self-hosted. Telephony, interruption handling, and retrieval are native to the system, proven and refined through enterprise deployments, not bolted on.”
Walter Badoo, co-founder and managing partner at 4DX Ventures, said the team is building what his firm believes will become “the defining voice infrastructure layer for the next billion users.”
Scaling up — and a note of caution
The funding will be used to scale enterprise deployments, expand engineering and go-to-market teams, and deepen product coverage across key regional markets. AethexAI currently has a team of 10 and expects to double headcount by the end of 2026.
It’s worth noting that the voice AI space for emerging markets has seen other attempts that didn’t quite stick. Reliable telephony integration at scale is hard, and localising models for dozens of languages and dialects — from Swahili to Arabic to Hausa — requires sustained investment. The startup’s approach of owning the entire stack gives it more control, but also means more surface area to maintain.
The company has real production deployments already, according to its investors.
Whether those deployments hold up as the customer base grows will be the real test. For now, the founders are betting that building specifically for the constraints of African and Middle Eastern networks — rather than adapting Western solutions — is the right bet.