I'm Dr. Imed Bouazizi, a computer scientist who has spent his career deciding how the world moves rich media: how video is compressed, streamed, rendered, and increasingly how immersive and AI-driven experiences are delivered over modern networks. Today I serve as Director of Technical Standards at Qualcomm, based in Dallas, building on a PhD (Dr. rer. nat.) in computer science from RWTH Aachen University.
What I work on
My focus sits where media meets the network. Over two decades I've contributed to the standards that power everyday streaming and the next generation of immersive communication:
- AI & agentic AI: characterizing how AI and LLM services consume the network, including agentic AI over Model Context Protocol tools, feeding the 3GPP SA4 study on 6G media.
- Real-time communication: conversational, low-latency media over WebRTC, from voice agents to interactive immersive sessions.
- Avatars and ARF: avatar-based immersive AR calls and the MPEG Avatar Representation Format (ARF, ISO/IEC 23090-39), covering animation, blendshapes, and skeletal playback.
- 3GPP & 5G media delivery: 3GPP Release-17 and beyond, including split rendering over IMS.
- Immersive & scene description: MPEG-I Scene Description (ISO/IEC 23090), which integrates Khronos' glTF 2.0 to describe interactive XR scenes.
- MPEG transport & systems: MPEG Media Transport (MMT, ISO/IEC 23008-1), the successor to MPEG-2 TS used in next-generation hybrid broadcast systems such as ATSC 3.0.
- Adaptive streaming: rate adaptation and delivery for DASH and HTTP streaming, work that has become foundational to how video is served today.
Recent work: AI traffic on 6G
My current focus is understanding how AI becomes a first-class network workload. I lead the 6G Testbed & AI Traffic Characterization project in the 5G-MAG Reference Tools effort: an open-source framework that measures and analyzes how AI and LLM services behave under realistic network conditions, built to support 3GPP SA4's new 6G media study. It pairs a scenario engine (chat, agentic AI over Model Context Protocol tools, image generation, multimodal, video understanding, and real-time WebRTC/WebSocket) with a Linux network emulator and 60+ AI-specific metrics such as time-to-first-token and time-per-output-token. See it under projects.
Standards leadership
I'm an active contributor and editor across MPEG, 3GPP, Khronos, and the IETF. Much of this work is paired with open, runnable reference software. I believe a standard is only as good as the implementation that proves it, which is why the projects on this site are working prototypes of the specifications I help write.
Research impact
My academic and industry research spans multimedia streaming, multipath video delivery, forward error correction, and immersive media. It has been cited more than 5,500 times (h-index 33, i10-index 97), across 80+ publications and numerous granted patents. Earlier in my career I also published foundational work in mobile and ad-hoc networking.
Beyond standards
I like to build. Outside of formal standards work I prototype 3D Gaussian-splatting viewers, WebNN and WebCodecs experiments, MPEG-I scene renderers, real-time voice agents, and developer tooling, most of which lives on this site or on GitHub. This site is also where I write about the economics and geopolitics of AI on the blog.