evidoria

← Back to browse

Good practice Imported

inABLE & Google Live Transcribe — AI Live-Captioning Pilot for Deaf Pupils in Kenya

Kenya · Kiambu · See the Kenya profile

Evidence: Descriptive / self-reported Top 59% 47/100 · Ask Evidence Copilot about this practice

Kenyan NGO inABLE and Google piloted AI live captioning at Kambui Primary School for the Deaf, logging 700+ hours across 108 pupils, with self-reported gains and plans to expand to more Kenyan schools.

108+
Students reached (2025)
15
Teachers reached (2025)
700+ hours
Classroom hours of live-captioning use logged (2025)
4 classrooms + computer lab + common room
Classrooms/spaces equipped (2025)
inABLE & Google Live Transcribe — AI Live-Captioning Pilot for Deaf Pupils in Kenya

Details

Maturity
Pilot
Promoter
inABLE
Period
2025–ongoing pilot
Keywords
live captioning, speech-to-text AI, deaf education, assistive technology, Africa

Context

Deaf pupils in Kenyan schools are taught primarily through sign language, but fluency varies widely among both teachers and students, and abstract subjects such as science and mathematics are especially hard to convey without a shared technical vocabulary; teachers report having to repeat instructions or write everything on the board to keep pupils following along.

Objectives

Pilot AI-driven real-time captioning of teachers' speech to improve comprehension and participation for deaf pupils.

Activities

In 2025, Nairobi-based nonprofit inABLE partnered with Google to pilot real-time captioning at Kambui Primary School for the Deaf in Kiambu County. Four classrooms, a computer lab and a common room were equipped with a 65-inch screen, a microphone and a laptop running Google's Live Transcribe app, which converts teachers' speech into text displayed live on the screen.

Results

inABLE reports the pilot reached over 108 students and 15 teachers, logging more than 700 hours of classroom use, with teachers and students describing improved comprehension and participation and less need to repeat instructions.

Conclusions

These outcomes are self-reported by the project lead rather than independently evaluated, and inABLE's own account flags that reliable internet, quality microphones and continuous teacher support are all necessary preconditions that are far from guaranteed in most Kenyan schools. inABLE and Google describe plans to replicate the model elsewhere in Kenya and beyond, but as of this pilot it remains a single-school deployment.

Implementation

Indicative cost
Low (< €50k) — The pilot equipped four classrooms, a computer lab and a common room at one school with a 65-inch screen, microphone and laptop each running Google's Live Transcribe app; costs are not itemised in public reporting.
Time to results
Short (< 1 year) — Piloted in 2025 and still ongoing as of the source reviewed, with inABLE and Google describing plans to replicate the model elsewhere in Kenya and beyond.
Staffing & skills
inABLE programme staff, Google (technology partner, Live Transcribe app), Kambui Primary School for the Deaf teachers

Conditions for success

  • Reliable internet connectivity
  • Quality microphones for accurate speech-to-text capture
  • Continuous teacher support and buy-in
  • Hardware provisioning (65-inch screen, microphone, laptop) per classroom

Common failure modes

  • Outcomes are self-reported by the project lead with no independent evaluation
  • Reliable internet, quality microphones and continuous teacher support are preconditions inABLE itself notes are far from guaranteed in most Kenyan schools
  • Remains a single-school deployment as of the pilot; replication elsewhere not yet evidenced

Where it fits

Governance type
NGO-corporate partnership pilot
Scale
single school (4 classrooms, 108 students, 15 teachers)
Income level
low-income

Commonly funded by

Philanthropic / foundation funding

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful