Building the voice infrastructure for a multilingual world.
LOBAKA builds voice technology that lets every person hear what matters to them — in their own language, on their own phone, in any room. Founded in Algiers by Moïse B. Selph.
Four people side by side, four different things in their ears. That's the whole product.
Sound belongs to the room. The room imposes its language.
A mic, a PA system, a room. Sound belongs to the space, and the space imposes its language on everyone in it. The speaker talks, the speakers amplify, and those who don't understand that language tune out — or disengage.
The industrial solution exists: simultaneous interpretation booths, dedicated receivers, trained technicians. It works, but it's expensive, it requires staff, and it only fits large venues. Most events — local conferences, performances, assemblies, ceremonies — simply can't access it.
We start from the other end: every attendee already carries the best audio receiver ever invented — their phone. All they need is the right channel.
LOBAKA was born in Algiers.
It builds voice tools for multilingual spaces — the places where simultaneous interpretation doesn't yet exist because the cost makes it impossible.
Founded by Moïse B. Selph, LOBAKA starts from a simple observation: every attendee already has the best audio receiver ever invented in their pocket. All they need is the right channel.
Today LOBAKA builds two products: LOBAKA Sasa for live multilingual audio, and LOBAKA Sauti for voice AI in African languages. More will follow.
Two products. One mission.
LOBAKA Sasa
Multilingual audio for live experiences.
An audio infrastructure that lets multiple audiences follow the same event through different language channels, directly from their smartphones.
Discover Sasa → Product 02LOBAKA Sauti
AI that speaks African languages.
A voice agent built to understand and speak African languages, making voice interfaces accessible to African users.
Discover Sauti →Three steps.
The organizer
Open a session on LOBAKA Sasa, create the channels you need — commentary, translation, audio description — and display the QR code on a screen or sign. No registration required for attendees.
The voices
Each voice — commentator, translator, describer — speaks on their channel from a phone or the venue's mic. One person can feed multiple channels if needed.
The listeners
They scan the code, choose their channel, and listen through their earphones. The stream is live, with no perceptible delay. They can switch channels at any time.
Earphones required — wired or Bluetooth, the ones everyone already owns.
Wi-Fi or mobile data connection needed in the venue.
Data usage is clearly disclosed before the attendee joins a channel.
We're looking for rooms, not letters of intent.
Multilingual concert, international conference, official ceremony, street performance, citizen assembly — if your event brings together people who don't share a language, we want to be there. We're looking for rooms for free pilot sessions: no contract, no invoice, just a room and an audience.
Fallback: moise@moisebselph.com