I’m a PhD researcher at IMT Atlantique working at the intersection of distributed systems, storage systems, cloud computing, and operating systems.

My main areas of interest include systems programming, distributed systems, Linux internals, networking, cloud-native infrastructure, and AI infrastructure. I enjoy working close to the system and understanding how software interacts with the kernel, network, hardware, and distributed infrastructure. Most performance stories and most outages are written at that layer, and I find it endlessly interesting.

I primarily work with C, Go, Linux, eBPF, Kubernetes, Docker, and networking technologies. I’m particularly interested in high-performance systems, distributed infrastructure, and the infrastructure required to build and run large-scale AI workloads.

What I’m working on

My PhD research focuses on improving data movement efficiency in large-scale distributed infrastructures through novel storage and networking abstractions. In practice that means designing operating-system and middleware mechanisms that optimize how data travels between distributed applications and storage tiers: zero-copy techniques, caching strategies, and kernel-level optimizations. I evaluate the prototypes on Linux platforms in the Grid’5000 research data center.

I got hooked on this problem during a research internship at Inria’s STACK team. I built a zero-copy system for multi-tier architectures handling image and video workloads: by overriding Linux’s send()/recv() system calls and using Redis as a shared buffer, intermediate nodes exchanged only lightweight descriptors until the client actually needed the data. The result was a 30% reduction in network bandwidth and 20–50% fewer I/O operations. Watching redundant data movement disappear from a live system convinced me this was the layer I wanted to spend my career in.

Before research

I came to research from engineering. Over several years I built backend systems for companies in Algeria: a secure authentication flow and a real-time WebSocket tracking system at Attraxia, a Spring Boot platform at SellSabeel that turned “here’s a repo with a Dockerfile” into a running AWS deployment in under ten minutes, 4G/5G/IMS lab environments at LabLabee, and a library and HR system for American Corner Oran that saved the staff twenty hours of manual work a week. That background shapes how I do research: I care about systems that run, not just systems that publish.

Some of my open source projects are on the projects page, and I write about Linux internals, containers, and systems programming here on this blog (originally published on dev.to).

Elsewhere

You can find me on GitHub, LinkedIn, dev.to, and Twitter. There’s also my CV.

If you want to chat about AI infrastructure, distributed storage, Linux internals, or systems engineering, email me at wassim.boussebha@imt-atlantique.fr.