Toku has launched Makimoto, an open source conversational AI initiative built around an MIT-licensed orchestration framework for data sovereignty, composability and long-term transparency in Asia-Pacific.
Toku has launched Makimoto, an open-source initiative for conversational AI in Asia-Pacific, anchored by Makimoto Kawa, an MIT-licensed orchestration framework for production-grade transcription pipelines. Available from 1 July as a managed transcription API hosted in Singapore, Kawa is designed around swappable components, with open-sourced glue code, configuration and APIs forming a modular orchestration layer.
Positioned as an alternative to closed AI stacks, Toku said Makimoto will remain permissively licensed, with no plans to adopt source-available or Business Source License-style models. The roadmap includes progressively opening more pipeline layers, advancing toward full transparency, community modifiability and an open-core architecture.
The launch also targets a growing need for compliant AI infrastructure as data residency rules tighten across Singapore, Indonesia and Vietnam. “With Kawa, customer data is processed in country, not just stored there. Every stage of the pipeline runs within a single jurisdiction, so our customers always know exactly where their data is going,” said Thomas Laboulle, Founder and CEO at Toku.
Kawa launches with real-time and post-conversation transcription APIs built on a five-stage modular pipeline, where each layer can be tuned or swapped. “We are obsessed with the orchestration layer, because models and components turn over every few months, but the layer that keeps them swappable is what makes the whole stack durable,” Laboulle said.
Makimoto will also interoperate with regional open models including SEA-LION and MERaLiON, while its governance structure separates community and enterprise editions.
