Skip to content
You must be logged in to sponsor ayutaz

Become a sponsor to yousan

@ayutaz

yousan

ayutaz
Japan

Hi, I'm yousan (@ayutaz), an independent open-source developer working on speech AI.

My mission is to make speech AI more open, practical, and globally accessible—especially technologies and knowledge originating from Japan.

I build tools that turn speech research into practical software for developers across Python, C++, C#, Rust, Go, WebAssembly, Unity, browsers, and resource-constrained edge devices.

What I'm building

  • piper-plus — A cross-platform, MIT-licensed neural TTS toolkit with Japanese and multilingual support, available across Python, C++, C#, Rust, Go, WebAssembly, and more.

  • sanoTTS-jp — A 559K-parameter Japanese TTS system that runs in real time on ESP32-S3, including Japanese text processing and pitch-accent handling on-device.

  • Vokra (in development) — A speech-first inference runtime written in Rust, targeting TTS, ASR, speech-to-speech, voice conversion, speaker identification, and VAD across CPU, Metal, CUDA, and Vulkan.

  • Ralomi (in development) — A private experimental research project exploring fully offline, free-form Japanese speech recognition on microcontrollers and resource-constrained edge devices. Its technical feasibility and public release are not guaranteed.

Projects marked in development are experimental. Their scope, architecture, features, and release schedules may change as research and validation continue.

How sponsorship helps

Your sponsorship helps fund:

  • GPU compute for model training, evaluation, and experiments
  • AI-assisted development tools such as Codex
  • Testing across runtimes, platforms, browsers, and physical devices
  • Maintenance, releases, documentation, and reproducible benchmarks
  • English documentation that makes Japanese speech technologies accessible worldwide
  • More time for sustained open-source development

I share a short development update with sponsors approximately every two to three months, covering recent releases, experiments, lessons learned, and upcoming work.

Sponsors can optionally have their name, link, or logo displayed on this Sponsors page.

All sponsorships support my personal open-source work and are not connected to my employer. Sponsorship does not guarantee a specific feature or custom implementation unless explicitly included in a sponsorship tier.

If you believe speech AI should be more open, local, and accessible across languages and devices, your support will help me continue building and sharing these technologies.

More about my work: Portfolio

@ayutaz

Help fund recurring GPU experiments and AI development tools such as Codex, while creating more time to maintain open-source speech AI and bring Japanese voice technologies to developers worldwide.

Current sponsors 2

@Kazuhito00
Private Sponsor
Past sponsors 1
@KazukiKuriyama

Featured work

  1. ayutaz/piper-plus

    Multilingual neural TTS (6 languages: JA/EN/ZH/ES/FR/PT, code supports SV) — C++, C#, Rust, Go, Python, npm (WASM). VITS + Prosody, streaming, CUDA/CoreML/DirectML. pip install piper-plus | npm ins…

    Python 218
  2. ayutaz/vokra

    Speech-first inference runtime in Rust — TTS / ASR / speech-to-speech / VC / speaker ID / VAD. An ONNX Runtime alternative that loads GGUF & safetensors directly: zero external dependencies, C ABI …

    Rust 13
  3. ayutaz/sanoTTS-jp

    559 K パラメータの日本語 TTS を ESP32-S3 で実時間合成。漢字かな交じり文の形態素解析・アクセント推定まで端末内で走る（M5Stack CoreS3 実機で確認）。推論は依存ゼロの C99、ブラウザ demo あり。arXiv:2608.21378 sanoTTS の日本語 clean-room 再実装。⚠️ コードは MIT ですが、配布モデルの重みは MIT ではありま…

    Python 67
  4. ayutaz/uPiper

    Unity TTS plugin: Piper neural synthesis + pure C# G2P (6 languages: ja/en/zh/es/fr/pt) + Unity Inference Engine. Windows/Mac/Linux/Android/iOS/WebGL ready. High-quality voices for games & apps.

    C# 33

Learn more about sponsoring developers and organizations.

Report abuse

4% towards $300 per month goal

@Kazuhito00

Kazuhito00 and 1 other sponsor this goal

Select a tier

$ a month

Choose a custom amount.

$3 a month

Select

Community Supporter

Support ongoing open-source speech AI development.

  • GitHub Sponsor badge
  • Development updates every two to three months
  • My sincere thanks for supporting open speech technology

$10 a month

Select

Voice AI Supporter

Help fund development tools, testing, and documentation for open-source speech AI.

  • Everything in the Community Supporter tier
  • Optional name and profile link on this Sponsors page

$50 a month

Select

Sustaining Sponsor

Provide sustained support for GPU experiments, releases, and global documentation.

  • Everything in the previous tiers
  • Optional logo and link on this Sponsors page
  • Opportunity to submit questions for upcoming development reports

$100 a month

Select

Project Sponsor

Support continued development of open-source speech AI projects.

  • Everything in the previous tiers
  • One 30-minute online discussion every three months
  • Opportunity to provide feedback on the public roadmap

This tier does not include guaranteed feature development or implementation work.

$500 a month

Select

Organization Sponsor

For organizations supporting open-source speech AI and helping bring Japanese voice technologies to a global audience.

  • Organization logo, link, and short description on this Sponsors page
  • Development updates every two to three months
  • One online discussion of up to 60 minutes per month
  • Up to two additional hours per month for technical investigation, issue review, or roadmap feedback
  • A short summary of sponsor-specific research or investigation

Custom implementation and guaranteed feature delivery are not included unless separately agreed.