Zimbabwean Founders Raise $2 Million For Ocular AI Data Lab

A Zimbabwean-founded AI company is raising a more fundamental question about the next stage of model development: what happens when the public web no longer provides enough useful data to improve frontier systems?

Ocular AI, co-founded by Michael Moyo and Louis Murerwa, has raised US$2 million in pre-seed funding to build an Applied AI Data Research Lab. BusinessTech Africa reports that the round was led by Drive Capital, with participation from Y Combinator, Alumni Ventures, 1745 Ventures, Orange Collective, MyAsia VC and angel investors.

The company is focused on a part of the AI stack that is less visible than a new chatbot but increasingly important: the creation and evaluation of high-quality data. Ocular began with voice AI, collecting conversations in a way that preserves features often lost when speech is reduced to a transcript. Those features include accents, pauses, interruptions, overlapping speech, tone and timing.

That distinction matters for African users. A system can perform well on clean, standardised speech while struggling with the accents, code-switching and conversational rhythms people use in everyday life. Better representation in training and evaluation data does not solve every problem of language access, but it gives developers a more realistic way to measure where systems work and where they fail.

Ocular is also moving beyond voice. The company has introduced audiovisual data that captures two participants’ audio and video separately while keeping the streams synchronised. This allows researchers to examine not only what people say, but also how they react, listen and communicate through visual cues. The approach is aimed at systems that can see and hear the world rather than only respond to text.

The company says its customers include leading AI laboratories and Fortune 100 companies, while its expert network includes thousands of vetted specialists. Its planned research lab will support new datasets and evaluation tools, including the Converse benchmark series. The first benchmark, Converse-STT, is designed to test speech-to-text models on real conversations rather than idealised recordings.

For Africa, the significance is broader than the size of the round. The continent is often discussed as a source of underrepresented data or as a market for imported AI products. Ocular’s model points to a different role: African founders building infrastructure that helps the global AI industry understand human communication more accurately.

That opportunity also comes with responsibilities. Voice and video data are highly sensitive. Consent, compensation, privacy, provenance and the ability of participants to withdraw must be treated as part of the product, not as paperwork added later. The quality of an AI dataset is inseparable from the conditions under which it was collected.

Ocular’s funding will support the expansion of its research and data operations. The company’s next test will be whether it can turn specialist data collection and evaluation into durable infrastructure for models that work across accents, cultures and real-world settings. That is a technical challenge, but it is also a question about who gets represented in the systems that are becoming part of everyday life.

Sources: BusinessTech Africa and Ocular AI.