Django API Overview - DRF, django-ninja, django-bolt and django-modern-rest

In this video, we'll look at Django API options, exploring the best modern packages for API-driven development in Django. We will take a look at Django REST Framework, django-ninja, django-bolt, and django-modern-rest - looking at the pros/cons of each.

https://www.youtube.com/watch?v=AFYNEGtsjKI

When to use NotImplemented

When should you return NotImplemented from a dunder method? Why not return False or raise an exception instead?

https://www.pythonmorsels.com/when-to-use-notimplemented/

Tracing np.add, all the way down

The author traces a simple np.add(a, b) call through NumPy’s internals, from Python argument handling and dtype dispatch to iteration strategy and the final SIMD-optimized C loop. Along the way, the post explores ufunc overrides, type promotion, dispatch caching, memory layout, GIL release, and CPU-specific optimizations that determine how NumPy actually performs an addition.

https://blog.veitheller.de/numpy.html

KV, Prefix, Prompt and Semantic Caching in LLMs, clearly explained

The post explains the four caching layers in LLM stacks: KV cache, prefix caching, prompt caching, and semantic caching, covering what each stores, their trade-offs, interactions, and the most common issues that prevent reuse. It includes first-principles explanations, code demos (with transformers and vLLM-style logic), production pitfalls, and practical takeaways for reducing recomputa...

https://x.com/_avichawla/status/2093265776266637739

Nifty Django Feature: Use Index for Custom Migration Operations

Part of the Nifty Django features series: You can hack Django's Index class to define custom migration operations on your model.

https://www.better-simple.com/django/2026/09/02/nifty-feature-use-index-for-custom-migrations/

Building AI Agents in Pure Python - Beginner Course

Build a fully functioning AI agent from scratch in pure Python, without frameworks, third-party tools, or vibe coding. The course focuses on the core mechanics behind agents so you understand how they work under the hood, not just how to prompt existing tools.

https://www.youtube.com/watch?v=c9AnqCeyxbI

experiential

An open source model gateway that provides one control plane across closed, open-source, local, and custom models.

https://github.com/experientiallabs/experiential

index-tts / index-tts

An Industrial-Level Controllable and Efficient Zero-Shot Text-To-Speech System

https://github.com/index-tts/index-tts

Neocarta

Library built for generating semantic layer graphs for query routing, query generation and data discovery.

https://github.com/neo4j-labs/neocarta

Fine-Tuning SOTA Object Detection Models on Real-World Datasets

Learn how to use these models, how to fine-tune them on diverse, specialized datasets that look nothing like their training data, and how to evaluate the results – all within PyCharm.

https://blog.jetbrains.com/pycharm/2026/08/fine-tuning-sota-object-detection-models-on-real-world-datasets/

OpenExecutive

AI-powered virtual executive team, a single coherent executive persona backed by 8 specialist Claude agents (FastAPI + Next.js).

https://github.com/SenteLabsAI/OpenExecutive

K-Dense-AI / scientific-agent-skills

Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 190,000+ scientists worldwide. 165 ready-to-use validated skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Pi, Antigravity, and the open Agent Skills standard.

https://github.com/K-Dense-AI/scientific-agent-skills

Citry

Fully typed frontend framework for Python with server events and Alpine.js, inspired by Vue and Livewire.

https://github.com/citry-dev/citry

Ruff, mypy, pytest, and then what?

Structural quality in the age of AI-written Python. What the standard toolchain checks, what it does not, and what a real agent-written repository looks like when you measure it.

https://codescan.dev/blog/ruff-mypy-pytest-and-then-what

TPU Inference Externalization Full Steam Ahead

The article examines Google’s push to make TPUs a first-class platform for external LLM inference, including TorchTPU, vLLM, SGLang, and a rapidly maturing open software stack. Benchmarks TPUv7 Ironwood against NVIDIA B200/B300 and dives into the kernel, memory, networking, and serving optimizations behind its performance-per-dollar gains.

https://inferencex.semianalysis.com/blog/tpu-inferencex-full-steam

cloud-in-a-bottle

Deploy, use, and share web apps on a server you control. Your apps, data, and infrastructure stay yours.

https://github.com/cloud-in-a-bottle/cloud-in-a-bottle

Reading __dict__ once permanently deoptimizes attribute access

Since 3.11 attribute access has not been a dict lookup, and reading dict once takes the specialized path away for good

https://deadlovelll.github.io/2026-09-05-reading-dict-deoptimizes-attribute-access/

hip-agent

A minimal coding agent harness that fits in the prompt.

https://github.com/changjonathanc/hip-agent

Teaching NumPy's ufuncs new tricks

The post explains how NumPy’s universal functions (ufuncs) work in C and how their internals were upgraded to support multi-output reductions, enabling the long-requested np.minmax function. It also covers ongoing work on segmented reductions for ragged data and converting existing functions into generalized ufuncs (gufuncs) for subarray processing.

https://labs.quansight.org/blog/teaching-numpys-ufuncs-new-tricks