🛠️ 5 self-hosted tools for your data science stack#
This article reviews five open-source alternatives for building a data science workflow without relying so much on SaaS: notebooks, tracking, orchestration, versioning, and storage.
What stands out#
- 📓 JupyterLab as your own hub
- 📈 MLflow for experiments and models
- 🔄 Airflow for orchestration
- 📦 DVC for data and models
- ☁️ More control and sovereignty over the stack
The idea is to replace subscriptions with your own infrastructure and more control over data, cost, and reproducibility.
🪄 Quick explanation#
Think of it as building your own lab.
Instead of renting every tool, you set up your environment and tailor it to how you work.
👉 More control, more independence, more repeatability.
More information at the link 👇
Also published on LinkedIn.
