Endava (NYSE: DAVA), a global provider of digital transformation, has announced a strategic partnership with SideFX, the creators of Houdini, the 3D animation visual effects software. This step makes Endava’s proven synthetic data generation pipeline publicly accessible through Houdini Labs, empowering artificial intelligence (AI) and machine learning (ML) teams to accelerate the development of high-accuracy computer vision models.
By combining quality 3D assets, VFX-grade pipelines, and experienced technical artists, Endava has closed the gap between synthetic and real-world data, allowing AI models to generalize effectively in production environments. These technologies enable the generation of comprehensive datasets—including object detection annotations, semantic segmentation masks, depth maps, and multi-spectral imagery—with pixel-perfect ground truth that eliminates human annotation errors.
“AI models, especially in computer vision, are only as good as the data used to train them. Open-source datasets and pre-trained models often come with licensing limitations or fail to meet the level of precision needed in production. Synthetic datasets solve these issues beautifully. They provide pixel-perfect annotations, enabling extremely high accuracy far beyond what’s achievable through human labeling. More than that, synthetic data can be intentionally generated without bias, including in scenarios that would be unethical or impossible to reproduce with real data – such as humans in dangerous environments,” said Dan Danciu, SVP Intelligent Automation at Endava Romania.
The collaboration comes at a pivotal moment for the synthetic data industry, which is experiencing unprecedented growth. Valued at over 300 million euros in 2024, the market is expected to expand rapidly, with analysts predicting that nearly 60% of data used in AI and analytics projects will come from artificially generated sources.
Unlike competing platforms that focus primarily on automotive or robotics applications, this collaboration specifically addresses the diverse needs of industries requiring specialized visual data, from medical imaging to industrial inspection systems. The integration leverages Houdini’s procedural generation capabilities to create infinite variations of training scenarios, ensuring AI models encounter edge cases and diverse conditions that would be costly or dangerous to capture in real-world settings.
What further differentiates Endava’s approach is the integration of tools for analyzing differences between synthetic and real-world data, helping teams understand and close the performance gap between training and deployment environments. Additionally, Endava provides built-in explainability mechanisms that highlight the exact image areas or pixels contributing to model predictions – enabling greater transparency and trust in AI systems.
Endava’s AI Vision team has already applied these tools in high-impact projects across healthcare, insurance, manufacturing, and digital media. Notable examples include enhancing dental X-rays with machine learning to support proactive diagnosis and using synthetic data to train AI systems for analog gauge reading in industrial environments. Beyond cost efficiency, the approach has enabled breakthrough applications that were previously impossible, such as industrial systems that operate in hazardous environments where human inspection would be dangerous.
“We’re incredibly excited to see our cutting-edge synthetic data tools become publicly accessible to the global AI community through this strategic partnership with SideFX. These aren’t experimental assets; they represent the culmination of our efforts to close the critical gap between synthetic and real-world data, allowing AI models to generalize effectively in production environments. By generating comprehensive datasets with pixel-perfect ground truth, these tools eliminate human annotation errors and ensure unprecedented accuracy for computer vision. Now, more teams can skip lengthy setup phases and go straight to solving truly meaningful problems, from industrial inspection to medical imaging”, further explained Dan Danciu, SVP Intelligent Automation at Endava Romania.
This collaboration reflects both companies’ commitment to democratizing access to high-quality AI workflows. With Houdini Labs now offering Endava’s data generation tools, teams across industries and academic institutions can adopt a proven solution to accelerate computer vision development and improve model performance at scale. As organizations increasingly embrace AI-powered visual analysis to enhance their digital capabilities, Endava is positioned to support this transformation through innovative synthetic data solutions.
