-
How to Test a Python Framework: Global State, Metaclasses, Parametrize, and Edge Cases
Testing a metaclass isn’t like testing a normal function — it fires when Python reads a class definition, not when you call it. This final post in the data-contracts series covers everything: autouse fixtures for global state isolation, testing metaclass behaviour by its observable outcomes, parametrize for every change type, and the heuristic edge cases…
-
Breaking vs Safe: How the data-contracts Diff Engine Detects Schema Changes and Notifies Consumers
Not all schema changes are equal. A new field is safe — consumers ignore it. A renamed field is breaking — consumers expecting the old name will fail silently. This post walks through how SchemaDiff classifies every change, how the rename heuristic works, and how the NotificationBus alerts only the teams that need to know.
-
Inside the Engine: How a Python Metaclass Automatically Registers Data Schemas
When you inherit from ContractBase, three things happen automatically — registration, field extraction, and validation — before the next line of your code runs. This post explains exactly how: the MRO walk, why get_type_hints() beats __annotations__, closures, and the schema registry that ties it all together.
-
Data Contracts in Python: Auto-Registering Schemas, Breaking Change Detection, and Consumer Notifications
One field rename. 50 broken pipelines. No warning. Here’s how I built a Python framework that detects breaking schema changes automatically — before they reach production.
-
Building a Plugin-Based Validation Engine in Python
How to validate, introspect, and safely load user-defined Python code with a real plugin engine as the example.
-

Outlier GPT
Python Pypi OpenAI NumPy Pandas Scikit-Learn PyTorch Git Why read this? This project addresses a major gap in modern analytics: detecting anomalies is easy, but explaining them is slow, manual, and often inconclusive. Outlier‑GPT automates that missing step by using an agentic LLM to merge internal data context with external real‑world signals, producing structured, defensible…
-

Arbitrage Free Neural Network for Option Pricing
Python PyTorch NumPy Pandas Scikit-Learn Introduction This project was inspired by the paper “Accelerated American option pricing with deep neural networks” by Anderson and Ulrych published in 2023. Given the extremely competitive nature of market-making, the ability to quote option prices near-instantaneously is essential to avoid arbitrage. Traditional numerical methods are often too slow for…