This post is also available in:
עברית (Hebrew)
Artificial intelligence is taking the world by storm, and indeed, it promises incredible innovation. However, turning an idea into a promising model is often a long, fragmented, and unpredictable journey. From messy data to disconnected workflows and reproducibility challenges, AI developers today face challenges that traditional software developers have long since solved, or ones they never had to. And for many organizations, especially those handling sensitive or classified data, the stakes – and the complexity – are even higher.
That’s exactly where DagsHub comes in.
DagsHub is an end-to-end platform purpose-built for AI development teams. By centralizing tools and workflows into a single, integrated hub, DagsHub dramatically accelerates model development in three key areas: turning raw data into structured datasets, tracking experiments, and model & dataset sharing.
Data Curation and Annotation
DagsHub starts by transforming multimodal or unstructured data into structured datasets. It connects directly to where the data is already being stored, builds enriched metadata tables, and supports both manual and AI-assisted labeling. This makes it easy to create training-ready datasets from vast amounts of data – paving the way for better model training.
Experiment Tracking
For training, DagsHub can integrate with training platforms such as Run:AI, OpenShift AI or AWS Sagemaker. This way, users can move their newly created datasets to their preferred model training platform with just two lines of code. As training runs, DagsHub tracks experiments – logging parameters, metrics, model weights, and other artifacts necessary for understanding the experiment’s progress. This way, the system allows users to better understand what worked, and what didn’t, and why.
Model & Dataset Management and Sharing
One common issue large organizations face is having multiple teams unknowingly work on similar AI projects in parallel. This duplication of effort not only wastes valuable time and resources but also leads to inconsistent results. DagsHub solves this issue by making AI models & datasets easily accessible across teams via a “model & dataset catalog”, filtered based on user permissions and clearance levels. In addition, the platform supports managing multiple versions of the same model, making it easy to adapt models for different use cases or customers.
Deployment
DagsHub can be deployed in the cloud or on-premises, but what makes DagsHub even more powerful is its support for secure deployments. The platform can be installed on air-gapped systems and with Red Hat OpenShift environments, making it ideal for organizations working with classified or sensitive data – like government agencies or defense clients.
Ease of Use & Portability
On top of everything, DagsHub is also incredibly intuitive. Within the platform, all low-level formats are based on open-source tools, so users don’t need to learn an entirely new system. Instead, they can continue working with the tools they already know—such as Git for code management, DVC for data versioning, MLflow for experiment tracking and model management, and Label Studio for data annotation. This approach lowers the barrier to entry, ensures compatibility with existing workflows, and allows teams to get up and running quickly without sacrificing flexibility or control.
The Team
DagsHub was founded in 2020 by Dean Pleban (CEO) and Guy Smoilovsky (CTO), childhood friends with deep technical roots. Dean, who studied physics and computer science and specialized in quantum computing, and Guy, a veteran of Israel’s elite tech unit with experience in backend, ML, and data engineering, saw a clear gap: software teams had powerful tools to manage code, but AI teams were stuck piecing together fragmented solutions. DagsHub was their answer—a unified platform for modern AI development.
Today, with over 50,000 users and customers, including high-security organizations, DagsHub is proving that the right tools can turn AI potential into AI reality.
DagsHub is a graduate of the INNOFENSE Innovation Center operated by iHLS in collaboration with IMoD. This unique acceleration program removes entrance barriers to the technological ecosystem, turning startups into mature, leading companies while connecting them with relevant investors, which is designed to strengthen the links between the civilian and defense markets via the collaborative development of the technologies, thus advancing and improving their integration in both markets.
For additional information, contact [email protected]
