Performing arts, in every audience’s language.
Our patented Supertitle™ system is what happens when Human-Centric AI, Edge AI, and Security come together to inspire and move people — on stage, in real time, in four languages, entirely on-device. Performing arts is where this architecture earned the right to be taken seriously elsewhere.
Captioning that’s part of the stage, not bolted onto it.
Traditional subtitle systems mount fixed LED screens beside the stage, forcing audiences to look away from the performance. Supertitle™ uses AI-driven holographic projection to create floating captions that integrate seamlessly into the stage design — real-time, multilingual, and entirely on-device. Audiences who would have lost the show to language now keep it.
Supertitle™ continuously learns speech, vocabulary, and lexical patterns from rehearsals and live performance, with human-in-the-loop review to approve or reject each finding.
The same architecture is the foundation we extend into worship halls, classrooms, conference venues, and disaster-resilience field operations. Supertitle™ is the proof point already on stage; every other room where language is a barrier is where we take it next. But on this page, the stage is what matters — the moment a line lands in a language an audience member did not expect to hear it in.
Demonstrated on Taiwan’s premier stage.
Weiwuying — National Kaohsiung Center for the Arts
Demonstrated October 2025 and January 2026 at Taiwan’s premier national performing arts venue, providing multilingual real-time captioning for live productions.
Featured on Taiwan Public Television
During the production of Kamada March Finale ~ Gin’s Departure (蒲田行進曲完結編〜銀ちゃんが逝く), Supertitle™ was profiled in a PTS (公視) special report on AI captioning innovation in the performing arts.
M678964
Registered
3255721
Registered
The PTS feature report.
Taiwan Public Television’s coverage of Supertitle™ during Kamada March Finale ~ Gin’s Departure.
The system on paper.
- PatentsTaiwan utility-model patent M678964 · Japan utility-model registration 3255721 (both 2026–2035)
- InventorsSophia Huang (黃舒郁) & Joseph Huang (黃奕勳)
- LanguagesMandarin · Japanese · Korean · English
- Live TranslationFully on-device. Zero cloud dependency.
- Codebase8,700+ lines of Python across 19 modules
- Built withClaude Code (Anthropic)
- MediaTaiwan Public Television (PTS)
From the proscenium to every room where language is a barrier.
Supertitle™ was first proven in front of paying audiences. The same on-device, multilingual, self-learning architecture carries into every room with the same shape: many languages, one moment, and words that must not leave.
Faith & worship Roadmap
Sermons, services, and liturgy in each congregant’s language, designed to run on hardware the community owns. Nothing streamed, nothing stored off-site. In design for 2027.
Education Roadmap
Designed for lectures and classrooms where one cohort spans four native languages — real-time captions, private by architecture. In design for 2027.
Conferences & keynotes
Professional-grade live captioning for international conferences. On-device processing keeps your data in the room. Available on the same Supertitle™ system already demonstrated on stage.
Crisis & disaster communications Roadmap
Emergency briefings, evacuation instructions, and cross-agency coordination that stay multilingual when uplink is the first thing to go — the engineering direction for 2026–2027. Same architecture, hardened for the conditions where it matters most.
Disaster resilience →Civil Training Roadmap
Real-time multilingual captioning intended for the Civil Training curriculum, so a single cohort could include native speakers of four languages without breaking a session. In design for 2026–2027.
See Civil Training →Mission-critical operations
For institutions whose conversations can never leave the room: designed for fully air-gapped deployment — no uplink, no telemetry — sovereign by design.
Sovereign by architecture →Frequently asked questions
What is the Supertitle system?
什麼是 Supertitle 動態字幕系統？
The Ariel Dynamic Supertitle System™ is a patented AI-powered holographic captioning system. It is protected by Taiwan utility-model patent M678964 (2026–2035) and Japan utility-model registration 3255721 (2026–2035), filed under the title Stage Emotion-Immersive Smart Subtitle Apparatus (舞台情感没入型スマート字幕装置).
What languages does Supertitle support?
Supertitle 支援哪些語言？
Supertitle supports Mandarin, Japanese, Korean, and English. All translation runs on-device, with zero cloud dependency.
Does Supertitle require an internet connection?
Supertitle 需要網路連線嗎？
No. Supertitle runs entirely on-device. Low, sub-100ms latency and zero cloud dependency — the same property that makes it the right base for field deployment anywhere connectivity cannot be assumed.
Where has Supertitle been demonstrated?
Supertitle 曾在哪些舞台展演過？
Weiwuying National Kaohsiung Center for the Arts. Featured on Taiwan Public Television (PTS) during Kamada March Finale ~ Gin’s Departure (蒲田行進曲完結編〜銀ちゃんが逝く).