Your MLOps platform, hosted in France
The reference platform for the machine learning lifecycle powered by MLflow, installed and maintained by DINAO. Your experiments, models, and training data remain on our French servers — without relying on an American ML cloud.
What is MLflow?
MLflow is a leading open source platform for managing the complete machine learning lifecycle. It covers experiment tracking, code reproducibility, model registry, and deployment, and has established itself as a MLOps standard.
Its Tracking component records parameters, metrics, and artifacts for every run, enabling you to compare training runs and identify the best model. The Model Registry versioned models and orchestrates their promotion across stages (staging, production), while MLflow Projects ensures reproducible packaging. MLflow also supports tracking and evaluation of LLM / GenAI applications.
Independent of libraries, MLflow integrates with scikit-learn, PyTorch, TensorFlow, XGBoost, and many others, via Python, R, Java SDKs and a REST API. With a PostgreSQL backend and S3-compatible artifact storage, it deploys in a container and scales naturally for data teams.
Host MLflow at DINAO
Resource tiers compatible with MLflow prerequisites (minimum 1 vCPU / 1 Go / 5 Go disque). Hosted in France, fully managed.
- 1 dedicated vCPU
- 2 Go RAM
- 20 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 2 dedicated vCPU
- 4 Go RAM
- 40 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 4 dedicated vCPU
- 8 Go RAM
- 80 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 8 dedicated vCPU
- 16 Go RAM
- 160 GB NVMe
- Daily backups
- Managed & monitored by DINAO
This application uses AI
The container hosts the application, not the AI engine (which requires dedicated GPUs) : inference is handled externally, with your own provider key. Prioritize a sovereign engine — Mistral AI, NumSpot, Scaleway, or OVHcloud AI Endpoints (France, GDPR) ; an international provider (OpenAI, Anthropic…) only if a specific capability requires it. Inference subscriptions are not included in hosting.
Technical details
You might be wondering…
What is MLflow used for?
MLflow manages the entire lifecycle of your machine learning models: experiment tracking (parameters, metrics, artifacts), code packaging, versioned model registry, and deployment. It is the backbone of an MLOps approach.
Do my datasets leave my servers?
No. On your DINAO instance hosted in France, your training data, experiments, and models remain on our servers. No data is transmitted to a third-party cloud.
Which ML frameworks are compatible?
MLflow integrates with most libraries: scikit-learn, PyTorch, TensorFlow, XGBoost, and many others, via Python, R, Java SDKs and the REST API. LLM/GenAI tracking is also supported.
Are technical skills required?
MLflow is designed for data/ML teams. DINAO handles installation, the backend (PostgreSQL, object storage), security, and updates; your data scientists focus on their models.
Can I change plans or export my data?
Yes. You can upgrade or downgrade at any time, and your experiments, models, and artifacts remain exportable — no vendor lock-in, as MLflow is open source.
