About
Thinking Machines is an independent directory of AI startups, institutions, and initiatives across Africa. The project is run by Opemipo Aikomo, Osarumen Osamuyi, and Precious Amuta through The Subtext.
Methodology
The directory is curated, and every entry is reviewed before publication. We add and update records when available evidence meets the criteria below.
Company classification
We consider any company that has a website and positions itself as an AI company. We group them into these categories:
- AI-native companies depend on AI for their core product or service.
- AI Infrastructure companies provide models, APIs, data, compute, or tools used to build and run AI systems.
- AI-enabled companies use AI as a material feature, while their underlying business could still operate without it.
- AI Services companies help other organisations adopt, implement, or customise AI.
- Non-AI companies do not have AI as a material part of the product customers use.
Inclusion and verification
We review source material for every addition. For companies, each entry needs evidence of an operational base or customer market in an African country; African founders alone are not enough. For institutions, programmes, and research initiatives, we look for a documented African presence, partnership, or work designed for African users and communities.
Keeping the directory current
New leads come from source directories, public reporting, organisation websites, research publications, and community referrals. An AI agent also monitors for potential additions; every proposed record is reviewed before publication.
Key resources
- AI Startups in Africa, compiled by Chinasa T. Okolo, provided the original company dataset. We reviewed those entries individually and continue to update their records.
- Propel Africa’s AI Focus directory is a major discovery resource for African startups building with AI. Its profiles help identify companies, markets, and product descriptions, which we check against company and other primary sources where possible.
- Organisation websites, research papers, model cards, public datasets, and reporting provide the evidence for institutional and research entries. Each record links to the material used for that entry.
Credits
The directory data is licensed under CC BY-NC 4.0. You may share and adapt it for non-commercial purposes with credit to The Subtext.
Opemipo Aikomo
Direction, production
Osarumen Osamuyi
Editorial
Precious Amuta
Data and curation
See the license notice.
Get in touch
For all enquiries, reach out to: