Rips through CPU-heavy file jobs on all cores via worker_threads. Zero deps, zero config. CLI + API

Python concurrency models comparison

Patches asyncio to add to_process — offload CPU-bound work to a separate process, get back an awaitable ProcessTask with full metadata, graceful cancellation, and first-class type support.

This repository contains a Retrieval-Augmented Generation (RAG) framework developed in C++ for high performance and scalability, with CPU/CUDA support for computational acceleration.

Scratchpad. This repository contains a Retrieval-Augmented Generation (RAG) framework developed in C++ for high performance and scalability, with CUDA support for computational acceleration.