Bridging the Gap at SummerSoc 2023, Creete, Greece 🇬🇷
Table of Contents
This summer, I had the pleasure of presenting our work, “Empowering Machine Learning Development with Service-Oriented Computing Principles,” at SummerSoC 2023 in beautiful Crete, Greece.
About SummerSoC 2023
SummerSoC is a unique symposium and summer school focused on service-oriented computing in its broadest sense. It brings together researchers and practitioners from diverse fields like cloud computing, IoT, service meshes, big data, and more. With participants from academia and industry worldwide, this event fosters learning, discussions, and collaborations.
Our Contribution
In our paper, we address the gap between machine learning development and service-oriented computing (SOC) principles. While SOC has transformed software development, ML development still struggles with challenges like managing diverse artifacts (e.g., models, code, and data) and orchestrating complex systems.
Our proposed framework, Everything as a Module (XaaM), introduces SOC principles into ML development by:
- Treating all ML artifacts as modules (models, code, data, configurations).
- Offering an architecture that enables efficient management and orchestration of complex ML systems.
- Leveraging SOC best practices to improve development maturity, streamline processes, and accelerate the creation of ML applications.
Why It Matters
By integrating SOC principles, ML development can achieve:
- Greater modularity for handling evolving systems.
- Enhanced efficiency in managing ML pipelines.
- Better scalability and maturity in building production-ready ML solutions.
Looking Ahead
We’re excited to continue advancing XaaM and exploring its potential to transform ML development.
A huge thank you to SummerSoC for organizing such a fantastic event and to everyone who engaged with us during our presentation!
Let’s continue the conversation! Feel free to connect or message me if you’d like to collaborate or learn more about XaaM.
References
The paper is published by Springer and available in the symposium proceedings:
📄 Empowering Machine Learning Development with Service-Oriented Computing Principles
Authors: Mostafa Hadadian, Viktoriya Degeler, Alexander Lazovik
