​Description

​What does it take to make AI feel immediate on real hardware? Join Tokyo AI (TAI) for a practical technical evening on portable inference, responsive AI demos and on-device applications.

​Designed for ML engineers, systems developers and people building edge AI products, the session will explore how runtime choices, memory constraints and application design shape end-to-end latency.

​We are planning three technical talks covering:

• ​Portable inference with ZML, from model execution to deployment across hardware.

• ​The engineering behind real-time AI demos, from edge robotics to interactive video.

• ​Building responsive on-device applications with local inference.

​When: Wednesday, December 16, 2026, 18:00-21:00 JST. This is a placeholder date and schedule, subject to confirmation.
Where: Tokyo, venue to be announced.

​Agenda

​Provisional schedule. Each 30-minute talk includes Q&A. Speakers, final titles and demonstrations will be announced once confirmed.

• ​18:00 - Doors open

• ​18:25-18:30 - Welcome

• ​18:30-19:00 - Talk 1: Portable low-latency inference with ZML (speaker to be announced)

• ​19:00-19:30 - Talk 2: Engineering real-time AI demos (speaker to be announced)

• ​19:30-20:00 - Talk 3: From local inference to responsive applications (speaker to be announced)

• ​20:00-21:00 - Networking

• ​21:00 - Doors close

​Talks

​These are tentative topic directions for discussion with the speakers, not confirmed talk abstracts.

​Talk 1: Portable low-latency inference with ZML: from model to local hardware

​A proposed practical look at taking a model from experimentation to local inference using ZML. Topics could include compilation and runtime choices, hardware portability, memory use and measuring latency across the full application. The final scope and hardware demonstration will be agreed with the speaker.

​Speaker: To be announced.

​Talk 2: Engineering real-time AI demos: from TORQ to interactive video

​A proposed walkthrough of the systems engineering behind TORQ and an interactive video-generation project. Possible topics include GPU memory bottlenecks, coordinating inference with a responsive application, and deciding which parts run locally or remotely. The emphasis would be on practical integration lessons and measured latency, with the demo format and final scope to be confirmed.

​Speaker: To be announced.

​Talk 3: From local inference to responsive applications

​A proposed talk on integrating on-device inference into an application people can use: model packaging, runtime integration, resource limits and the tradeoffs between offline capability and responsiveness. Speaker and demonstration to be confirmed.

​Speaker: To be announced.

​Organizers

​Ilya Kulyatin is an entrepreneur with work and academic experience in the US, Netherlands, Singapore, UK, and Japan. He holds a BA in Economics, an MA in Finance, and an MSc in Machine Learning. He's a 3x founder, now helping Japan grow the local AI ecosystem through a not-for-profit community, Tokyo AI (TAI), while building an AI-native system integrator and solutions provider, Foundry Labs株式会社.

​Supporters

​Foundry Labs K.K. is a Tokyo-based AI systems integrator and solutions provider, delivering end-to-end support for enterprises: from strategy design through implementation, deployment, and operations. They tailor AI to each client's operational, regulatory, and security requirements, with hands-on experience across finance, government, and industry, and a track record of shipping production systems in secure and regulated environments.

​About TAI

​Tokyo AI (TAI) is the largest international AI community in Japan, with 5,000+ members mainly based in Tokyo: engineers, researchers, investors, product managers, and corporate innovation leaders. Through 80+ events a year and 300+ speakers spanning startups, enterprises, and academia, TAI connects the people building AI in Japan with the global ecosystem, working to transform Tokyo into a global AI hub.

​Privacy Policy

​We will process your email address for the purposes of event-related communications and ongoing newsletter communications. You may unsubscribe from the newsletter at any time. Further details on how we process personal data are available in our Privacy Policy.