Explore the directory

An example project that builds a question-answering system over a collection of videos, extracting frames and transcripts, embedding them into a shared multimodal space, and answering questions with a vision-language model.

• RAG
• Tutorials

A set of components for building an AI gateway on top of the Praxis proxy, adding provider-aware routing, protocol translation, response storage, guardrails, and observability at the proxy layer.

• LLM gateways

The code generator for Soto, an AWS SDK for Swift, turning AWS Smithy service models into Swift client code, usable as a Swift Package Manager build plugin so a project only compiles the AWS services it actually uses.

• Libraries
• APIs

A ggml-based C++ implementation of Stable Audio 3 and the Stable Audio Tools model family for cross-platform and embedded use, providing generation and LoRA training CLIs, an embeddable C ABI, and a local browser studio.

• Generative media

The source for Crowdin's Knowledge Base and Developer Portal, an Astro and Starlight site documenting Crowdin's localization platform and its API for developers integrating translation management into their software.

• Documentation tools

A benchmark suite that evaluates typed decision models on more than raw accuracy, measuring probability calibration, selective-risk coverage, latency and reproducibility, comparing TypeSafe Jev against a zero-shot classification baseline.

• AI evals

An on-device context and insights SDK for Swift apps that reads real-world signals on the device, without collecting personally identifiable information, to help an app respond to a user's real-world context.

• Mobile
• Libraries

A free, open-source IntelliJ plugin that provides AI pair programming through Editor Code Assistant (ECA), connecting to an external eca server process for chat, code suggestions and context management inside the IDE.

• Coding agents
• Editors

A lightweight, embeddable vector database and multimodal AI memory engine built for edge devices such as Raspberry Pi and NVIDIA Jetson, combining vector search, graph relationships, and offline RAG in one process.

• Agent memory
• Databases

A community-maintained, async Scala client for the Pinecone vector database API, covering vector, collection, index, inference, and assistant operations with minimal dependencies.

• Databases

A Python library and toolkit for GraphRAG built entirely on plain PostgreSQL, combining vector search, full-text search, and knowledge-graph traversal in a single SQL query without Neo4j or a separate vector database.

• RAG
• Databases

The official MCP server for Zenrows' web-scraping infrastructure, giving AI agents and assistants reliable, real-time access to live web pages, including sites protected by anti-bot measures.

• MCP servers
• Browser automation

A CUDA kernel implementation that runs the Qwen3.6-27B language model on four Tesla V100 GPUs at up to 366 tokens per second, using hand-written NVFP4 kernels and speculative decoding.

• Inference

A static security scanner that grades MCP tool definitions A through F for risks such as prompt injection, data exfiltration, privilege escalation, and supply-chain backdoors, before an agent is allowed to call them.

• MCP servers
• Security

A portable skill that gives coding agents a senior-engineer standard for writing code across multiple languages and frameworks, covering architecture, testing, security, performance, and validation.

• Agent skills

A network-faithful simulator that predicts the serving performance of large LLM deployments before you buy or reserve hardware, modeling the fabric at packet level rather than as a simple bandwidth estimate.

• Inference

An AI agent framework written in Java, maintained by a small named group of contributors.

• Agent frameworks

A vector compression toolkit for AI workloads that uses TurboQuant codecs to shrink embeddings and KV-cache tensors roughly tenfold, with typed encode and decode contracts and measured byte accounting.

• RAG
• Inference

An open-source Apache Spark extension for large-scale document processing that combines OCR, LLMs, and vision-language models to extract structured data from PDFs and images at scale.

• RAG
• Data engineering

A Python framework for building multi-agent AI systems around a guild model, where pluggable agents share memory and communicate to perform long-running tasks such as software development or research.

• Agent frameworks

A retrieval-only question-answering system for documentation that makes zero LLM calls at query time, matching questions against pre-extracted answer blocks using local vector search instead of generation.

• RAG

A production-style Retrieval-Augmented Generation pipeline built from scratch without LangChain, combining FAISS and BM25 hybrid retrieval with cross-encoder reranking and a built-in evaluation harness.

• RAG
• AI evals

A small RAG application that answers questions about uploaded PDFs or text files by first searching the documents for relevant passages, then grounding an OpenAI GPT-4o answer in that retrieved text.

• RAG

A reproducible serving configuration, vLLM patches, and measured benchmarks for running the Qwen3.8-27B model on one or two RTX 5090 GPUs at 262K context for concurrent coding agents.

• Inference