About Speculum AI
The AI Runtime Company
Models will change. Runtime stability will remain.
Speculum AI builds the independent runtime layer for production AI: Mirror OS for AI runtime stability and Mirror Token for AI runtime efficiency.
Ecosystem · Programmes · Partners
Our story
The architecture was ready before the market asked for it.
Built in research first. In market since January 2026.
Speculum AI was founded on one conviction: as AI moves from answering questions to taking actions, the systems that govern it must stand outside it. Intelligence does not define its own authority.
We built the runtime architecture first, then took it to market — which is why one architecture can extend across AI domains while several markets open in parallel.
Timeline
From stealth research to market.
- January 2024
Stealth research begins
Research begins in Taiwan on an independent runtime architecture for production AI.
- October 2025
Emerging from stealth
Speculum AI moves from research into market preparation, bringing its runtime architecture into enterprise and ecosystem conversations.
- January 2026
Go-to-market begins
Commercialisation begins, with Japan as the first focus for enterprise engagement and market development.
- Q4 2026
Customer engagement and business expansion
Customer engagement deepens and business expansion advances across Taiwan, Japan and Singapore. Speculum AI confirms plans to establish business entities in Tokyo and Singapore.
- Today
Finalising for release
Mirror OS is in active market engagement and completing final product alignment with our distribution partners ahead of its official release targeted for Q1 2027. Mirror Token is in final productisation with our platform partners, with availability targeted for Q1 2027.
Leadership
Leadership.
Company-building, AI systems research, security and standards — brought together to make production AI stable, trustworthy and controllable at runtime.
Dominique Tu
France / Asia
Founder & CEO · Chief Scientist
- Serial entrepreneur with four exits
- Built an IP portfolio spanning three patent families and five applications, including Taiwan utility model M686084, granted in August 2026
- Executive experience across banking, insurance and cybersecurity
- Leadership roles across Standard Chartered Bank, Nokia, PalmSource, IntelliSync and Kii
Dr. Hen-Hsen (HH) Huang
Taiwan
AI Model & Systems Advisor
- Research Lead / CTO, Taiwan Sovereign AI
- Assistant Professor, National Chengchi University
- AI systems researcher with 2,000+ citations
Dr. Meng-Chow Kang
Singapore
Founding Security & Standards Advisor
- Associate Professor, HKUST
- Former Amazon APAC Principal, Security Assurance
- Former Microsoft Greater China Chief Security Advisor
- Former Cyber Security Advisor to the Monetary Authority of Singapore (MAS)
Designated Managing Director, Singapore
Government & Institutional Relations
Upon establishment of the Singapore entity
Where we operate
Built in Asia. Scaling where trusted AI matters most.
Our foundation is in Taiwan. We are building long-term operating presence in Japan and Singapore — combining deep-technology R&D, enterprise market development and Southeast Asian reach with a partner-led path to global scale.
Taiwan
Founding, R&D & IP base
The centre of our core runtime research, product engineering and intellectual-property development.
Japan
Lead commercial market
Our first focus for enterprise engagement, distribution alignment and long-term market development.
Singapore
Southeast Asia expansion
Building our regional operating presence for business development, institutional relationships and expansion across Southeast Asia.
Global
Partner-led expansion
Developing the ecosystem, channel and investor relationships that can support selective entry into Europe and the United States.
Careers
Help make the world’s AI more stable and trustworthy.
AI is moving from experimentation into production. The next chapter will be defined not only by what AI can do, but by whether organisations can trust it while it is running.
Speculum AI is building the independent runtime technology for that future. If you want to help define a new category and bring it to global markets, send us your résumé. We want to hear from exceptional people across engineering, AI and data science, and go-to-market.