Pinned Loading

1. Mini Autograd Mini Autograd

   1

   """

   2

   A minimal reverse-mode autograd engine, built from first principles.

   3

   No dependencies beyond the standard library.

   4

   5

   Core idea:

2. Mini Transformer Mini Transformer

   1

   """

   2

   A minimal Transformer (decoder-only, GPT-style), built from first principles

   3

   on top of a scalar reverse-mode autograd engine. No dependencies.

   4

   5

   Essence of the design:

3. databricks-churnml databricks-churnml Public

   End-to-end SaaS churn prediction platform on Databricks using Medallion architecture, MLflow, and deployed via Databricks Asset Bundles

   Python

4. sales_lakehouse sales_lakehouse Public

   Enterprise Data Engineering with Azure, Databricks, PySpark and modern Lakehouse architecture.

   Python

5. azure_de azure_de Public

   End-to-end Azure data engineering pipeline transforming sales data into Power BI-ready datasets using a medallion architecture.

   Jupyter Notebook