メインコンテンツへスキップ
見出し画像

Pushing the Limits of Edge AI: Antoine Sauvage (EDGEMATRIX) Takes the Stage at Tokyo AI

    How do you make AI respond quickly and reliably when compute, memory, and power are strictly constrained?

    Deploying complex models in real-world physical environments is fundamentally different from running inferences in high-powered cloud data centers. In the field, hardware limits are absolute, connections drop, and conditions change without warning.

    To tackle these challenges, the upcoming Tokyo AI (TAI) meetup convenes engineers and researchers working close to the hardware for an evening dedicated to efficient inference. Highlighting the lineup is Antoine Sauvage, Deputy Director of Engineering at EDGEMATRIX, sharing hard-earned field insights on running reliable vision models on embedded GPUs.

    Featured Speaker: Antoine Sauvage (EDGEMATRIX)

    Talk 1: From Model to Field: Engineering Reliable Edge AI Applications on Embedded GPUs

    18:30 – 19:00 (Includes Q&A)

    Antoine Sauvage brings deep engineering experience from EDGEMATRIX, a leader in edge AI platforms and infrastructure. In this session, Antoine will cut through theoretical optimization and dive straight into the operational realities of building real-time computer-vision systems on NVIDIA Jetson platforms.

    Key Takeaways:

    • GPU-Accelerated Video Pipelines: Balancing high-throughput processing against constrained compute and thermal budgets.

    • Tackling Edge Volatility: Architecting systems that gracefully handle unreliable cameras, noisy video feeds, and intermittent network connectivity.

    • Production Operations: Best practices for remote device management, telemetric monitoring, and OTA updates required to maintain 24/7 reliability in mission-critical deployments.

    画像

    📅 Date: Friday, 13 November 2026
    ⏰ Time: 18:00-21:00 JST
    📍 Location: Shibuya, Tokyo
    🔗 Register: https://luma.com/xkb4ai5f

    This event is made possible through the support of:

    • Foundry Labs K.K.: Tokyo-based enterprise AI systems integrator specializing in secure, mission-critical deployments in regulated environments.

    • Plug and Play: Global venture capital and innovation platform backing tech founders and domestic startups across Japan.

    あなたへのおすすめ