Open Source Experience 2025#

Conference Overview#

In December 2025, DataLab 1.0 will be presented at Open Source Experience 2025 in Paris, one of Europe’s largest events dedicated to open-source software and communities. This presentation marks the official announcement of DataLab’s evolution into a modular platform.

Event Details:

Date:: December 11, 2025
Time:: 14:15 - 14:35
Location:: Cité des Sciences, Paris
Room:: Salle Plénière (Plenary Hall)
Track:: Artificial Intelligence and Scientific Computing
Format:: Talk (20 minutes)

Note

Topic Change: Due to patent process timing constraints at CEA, the presentation topic was changed from the originally planned X-ray scene reconstruction case study to a broader overview of DataLab 1.0. Information about the original topic is preserved below for reference.

Presentation#

Title (French): DataLab 1.0 : Une application complète et une plateforme modulaire pour le traitement de données scientifiques

Title (English): DataLab 1.0: A Complete Application and a Modular Platform for Scientific Data Processing

Speaker: Pierre Raybaut - Executive VP, Engineering, CODRA

Abstract#

DataLab 1.0 marks a major milestone in the evolution of the open-source scientific application originally designed by CODRA. This release consolidates two complementary dimensions:

a complete graphical application for exploring, processing and analyzing 1D/2D scientific data;
a modular platform whose core processing components are now available as reusable libraries.

At the heart of this architectural shift lies a new object-oriented processing library, designed as part of the DataLab Core Architecture Redesign project funded by the NLnet Foundation. This library enables developers and domain experts to easily create their own data processing tools while reusing the same core components that power DataLab.

The talk will highlight the key features of DataLab 1.0, including its redesigned architecture and new user-facing functionalities, and will demonstrate how it can serve both as a ready-to-use application and as a foundation for building domain-specific tools in scientific and industrial contexts.

Original Planned Topic#

The presentation was originally planned to focus on an advanced industrial use case developed by CODRA for CEA.

Original Title: Automatic Reconstruction of X-ray Scenes with Python and DataLab

Original Abstract:

In the field of non-destructive testing, the CEA entrusted CODRA with the specification, design, and development of software for the automatic reconstruction of radiographic images, entirely based on open-source components and contributing to the scientific Python ecosystem.

Key Technical Aspects:

Geometric Pattern-Based Reconstruction: Using a grid of geometric markers placed within the radiographed field, the algorithm automatically assembles a complete scene from partial images, without any prior external information (position, orientation, magnification).
Complex Parametrizable Processing: The complex and configurable processing workflow was developed and refined using the open-source DataLab platform, remotely controlled to dynamically visualize each step of the pipeline.
Real-World Application: The creation of this software paves the way for new applications in the fields of industrial imaging and security.

DataLab 1.0: Key Features and Architecture#

This presentation will showcase the major advancements in DataLab 1.0:

Modular Architecture

The new architecture separates concerns into distinct, reusable components:

Sigima: Core object-oriented processing library for signals and images
DataLab GUI: Complete application built on Qt/PlotPyStack
Reusable Components: Libraries that can be integrated into custom applications

NLnet Foundation Support

The DataLab Core Architecture Redesign project, funded by the NLnet Foundation, enabled:

Complete refactoring of the processing engine
Creation of the Sigima library for standalone use
Improved modularity and extensibility
Better separation between computation and GUI

Dual Purpose Platform

DataLab 1.0 serves two complementary roles:

Ready-to-Use Application: Full-featured GUI for data exploration, processing, and analysis
Development Platform: Foundation for building domain-specific tools

New Capabilities

Version 1.0 introduces significant improvements across all areas:

Interactive Editing: Modify creation and processing parameters after object creation, with automatic updates
Menu Reorganization: New “Create”, “ROI”, and structured menus for better workflow
Enhanced Analysis: Multi-object property editing, comprehensive result management, and automatic ROI analysis updates
Advanced Processing: New operations including pulse analysis, 2D resampling, coordinate transformations, and frequency domain filters
ROI & Annotations: Copy/paste, import/export, grid creation, and multi-object editing capabilities
Visualization: Optimized performance for large datasets, customizable result display, and DateTime support
File Formats: Support for FT-Lab formats, non-uniform coordinates, and enhanced HDF5 handling
Cross-Platform: Version-specific configurations allowing multiple major versions to coexist

Why This Matters for DataLab#

This presentation is significant for several reasons:

Major Release Milestone: DataLab 1.0 represents years of development and architectural refinement, marking the platform’s maturity.
NLnet Foundation Recognition: The project’s funding by the NLnet Foundation validates DataLab’s importance in the open-source scientific software ecosystem.
Platform Evolution: Shows how DataLab has evolved from a standalone application to a modular platform that can be integrated into custom solutions.
Community Growth: Demonstrates DataLab’s commitment to serving both end-users (through the GUI application) and developers (through reusable libraries).
Industrial and Academic Adoption: Highlights DataLab’s versatility in meeting the needs of diverse user communities.
CEA Partnership: Reinforces the strong relationship between DataLab and CEA, one of its major supporters.

Looking Forward#

This presentation at OSXP 2025 will:

Announce and showcase DataLab 1.0’s release and new architecture
Demonstrate how the modular design enables custom application development
Highlight the successful collaboration with the NLnet Foundation
Present concrete examples of how DataLab can be used both as an application and as a development platform
Inspire organizations to adopt DataLab for their scientific data processing needs
Foster community engagement and contributions to the DataLab ecosystem

DataLab 1.0 represents a significant evolution - from a specialized application to a comprehensive platform that empowers users to build their own data processing solutions while maintaining the simplicity and power of the complete GUI application.

The originally planned X-ray scene reconstruction project remains an excellent example of DataLab’s capabilities in advanced industrial applications, demonstrating how the platform bridges the gap between research and industrial applications while combining scientific rigor, industrial robustness, and open-source collaboration.

Note

Upcoming Event: This presentation will take place on December 11, 2025. Check back after the event for presentation materials, feedback, and potentially a video recording.

Conference Track#

The presentation is part of the Artificial Intelligence and Scientific Computing track, which is appropriate for DataLab 1.0’s positioning at the intersection of:

Modern scientific computing (signal and image processing)
Platform architecture and modular design
Open-source development methodologies
Industrial and academic applications

This track placement highlights DataLab’s evolution as both a complete scientific application and a foundation for building specialized tools in the scientific Python ecosystem.

Resources#

Open Source Experience 2025 website
Talk details and schedule to be published closer to the event