Text-to-speech
Turn written text into spoken narration for lessons, articles, and books.
Software engineering and Ethiopian-language voice AI.
We build and run Sagale, our voice platform for Amharic, Afaan Oromo, and Tigrinya. We also fine-tune existing voice models for Ethiopian languages using licensed and publicly available datasets, and publish the adapters and evaluations. Our client work includes system integration, cloud and DevOps, data pipelines, and applied AI.
What we build
We operate Sagale and provide software engineering services to clients.
Sagale capabilities
Sagale is for people preparing spoken Amharic, Afaan Oromo, or Tigrinya: creators, publishers, educators, media teams, and diaspora families.
Turn written text into spoken narration for lessons, articles, and books.
Make speech easier to search, edit, quote, and archive.
Prepare long-form text for listening, including book and document workflows.
Create Amharic, Afaan Oromo, or Tigrinya voice tracks for existing video and audio.
Open source
We fine-tuned Chatterbox Multilingual v3 for Amharic and Afaan Oromo using licensed datasets. The adapters, evaluations, and text-processing code are publicly available.
ኢትዮጵያ የረጅም ዘመን ታሪክና ኩሩ ባህል ያላት ጥንታዊት ሀገር ናት።
How we work
Projects include an agreed scope, regular working releases, documentation, and a clear maintenance or handoff plan.
Forms, portals, and admin screens designed for the people who use them every week.
Older systems, vendor APIs, data sources, and identity providers, connected and kept in sync.
Infrastructure, releases, logs, alerts, and the maintenance work that starts after launch.
Documentation and code another engineer can pick up. You are not locked to us.
Tell us what you need to build, replace, integrate, or repair.