WrenAI

Open-source GenBI: generative BI for AI agents.

https://github.com/Canner/WrenAI

In-House LLM Serving at Netflix

Netflix built an in-house LLM serving platform using vLLM and NVIDIA Triton, integrating self-hosted models into its existing production infrastructure through unified gRPC and OpenAI-compatible APIs. The article details production lessons around model packaging, version compatibility, zero-downtime deployments, observability, and scaling constrained decoding by moving bottlenecks from s...

https://netflixtechblog.com/in-house-llm-serving-at-netflix-a5a8e799ea2c

google / agents-cli

The CLI and skills that turn any coding assistant into an expert at creating, evaluating, and deploying AI agents on Google Cloud.

https://github.com/google/agents-cli

Reusable scenario based test framework

Write your tests once. Run them on every product variant, device or technology — without rewriting a single line.

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

Building Agentic Workflows in Python with LangGraph

In this article, you will learn how to build a complete agentic workflow in Python with LangGraph, from a single model call to a tool-using agent with persistent conversation memory.

https://machinelearningmastery.com/building-agentic-workflows-in-python-with-langgraph/

xarray-sql

An experiment to query Xarray datasets with SQL.

https://github.com/xqlsystems/xarray-sql

venv_manager

A powerful CLI tool for managing Python virtual environments with ease.

https://github.com/jacopobonomi/venv_manager

Django: introducing django-crawl

Adam Johnson introduces django-crawl, a tool that uses Django’s test client to crawl a site and uncover broken links, exceptions, and other hidden issues without making real HTTP requests. It can run as a management command or automated test, providing a fast safety net that found seven bugs even in a project with 100% test coverage.

https://adamj.eu/tech/2026/07/22/introducing-django-crawl/

Mage

A Lightweight, Research-Friendly Multimodal Model Family.

https://github.com/microsoft/Mage

LLVMLITE in the Browser

This notebook explores how LLVM's optimization pipeline transforms a simple loop, step by step, using llvmlite running entirely in the browser via WebAssembly. We follow one small C++ function through three distinct stages — from source to raw stack-based IR (Intermediate Representation), through SSA form, to a closed-form calculation. After each stage, we inspect the control flow graph ...

https://notebook.link/@anutosh491/llvmlite

tsauditor: a pre-flight check for time-series data before you trust your metrics

Checks a finished time-series dataframe for common data problems before training, like leaked future information, weird spikes or stuck values, missing days in the date range, and features that drift.

https://github.com/imann128/tsauditor

cosmos-framework

Our inference and training framework to run on the Cosmos Models.

https://github.com/NVIDIA/cosmos-framework

LoopGain

An open-source cost controller for AI agent loops.

https://github.com/loopgain-ai/loopgain

browser-use / video-use

Edit videos with coding agents

https://github.com/browser-use/video-use

Emily2040 / seedance-2.0

Comprehensive production pipeline for quad-modal AI filmmaking with Seedance 2.0

https://github.com/Emily2040/seedance-2.0

percentify

Exploratory stats and data-quality diagnostics for pandas and Polars DataFrames. One call each.

https://github.com/data-centt/percentify

How To Train a Generative Kick Drum Model on Your Old Linux Desktop With 6GB of VRAM

The author trained a text-conditioned generative kick-drum model from 13,615 audio samples on a seven-year-old GPU with only 6GB of VRAM, using a VAE, latent diffusion model, and HiFi-GAN vocoder. The project shows that useful generative models can be built with modest hardware and personal datasets by compressing data into latent space, choosing practical constraints, and using serverle...

https://www.zhinit.dev/blog/training-a-kick-drum-diffusion-model