Language, culture and contrasting experiences with Kabakoo’s AI Mentor

November 21, 2026
Kabakoo team members work on laptops at La Maison de l’Étonnement in Bamako in 2024. Context photograph.

Conversations with Kabakoo’s AI Mentor differed by language: entrepreneurship, endogenous knowledges and local culture appeared more often in Bamanankan, while digital and technical questions appeared more often in French. Alongside those patterns, learner accounts describe both encouragement and difficulty with the same product.

1. Bamanankan voice accounted for substantial use

Kabakoo’s Mentor works in French text and Bamanankan voice. Across 11,171 bilingual conversations, nearly 40% of voice interactions happened in Bamanankan. That figure describes voice interactions within the conversation data—not the share of learners choosing a language. Kabakoo describes its multilingual learning assistant in its overview of the Kabakoo AI System.

2. The topic mix differed between languages

In the bilingual conversation data, entrepreneurship, endogenous knowledges and local culture topics ran twice as often in Bamanankan as in French, at 10.5% against 5%. Digital and technical questions showed the reverse pattern: they came about four times more often in French than in Bamanankan, at 13% against 3%. These are patterns in conversation topics, not measured learning outcomes.

3. A rejected submission preceded one learner’s departure

In March 2026, a learner in Bamako recorded the introduction video required for her first challenge in Kabakoo’s upskilling program. The AI Mentor found that it did not meet the criteria and rejected it. She never tried again. She later explained: “I dropped out after the first challenge, when the AI Mentor did not accept my introduction video.” Focus groups and hundreds of individual phone calls with program dropouts also revealed revision loops: many learners reported remaking a video up to 5 times; some described 10 attempts on one submission without validation.

4. Maimouna experienced feedback as encouragement

Maimouna described a different encounter with the same AI Mentor. She said it was encouraging, gave her feedback, showed her what to add or change, and helped her find the courage to submit videos. Her account sits alongside the reported revision loops: the same product featured in accounts of both encouragement and repeated submissions without validation.

Kabakoo Dispatch

Keep up with what Kabakoo is building.

Research, field notes, opportunities and stories from across Kabakoo, delivered directly to your inbox.

Subscribe on Substack