$ show patchy631/ai-engineering-hub
This repository is a curated collection of 93+ production-ready AI engineering projects, spanning beginner to advanced levels. It serves as a practical resource for building AI applications, with a strong emphasis on agent orchestration (CrewAI, AutoGen, OpenAI Swarm), Retrieval-Augmented Generation (RAG) pipelines, Model Context Protocol (MCP) servers, and LLM evaluation frameworks. The dominant character is application development: most projects are deployable apps using Streamlit, Next.js, FastAPI, or as standalone MCP servers. Notable sub-projects include a full visual workflow builder (open-agent-builder) with Convex backend, a multi-modal RAG pipeline (notebook-lm-clone), and a comprehensive audio analysis toolkit. While primarily application-focused, the repository also contains data/research elements such as model fine-tuning scripts (GRPO, DPO, SFT), synthetic data generation (SDV), and LLM comparison dashboards with Opik/DeepEval evaluation. The codebase is heterogeneous in quality: some projects exhibit clean architecture and reusable patterns, while others are simple demos with boilerplate code and hardcoded dependencies. Overall, it is an excellent educational resource for AI engineers seeking practical, runnable examples, but not a single cohesive product.
The LLM identified 5 files worth highlighting for reuse:
A reusable agent executor that supports multiple LLM providers (Anthropic, OpenAI, Groq) with MCP tool resolution, variable substitution, and streaming. It demonstrates a clean abstraction for executing AI agents within a workflow engine, handling tool calls, state management, and error boundaries. Could be extracted as a standalone library for any workflow system.
A self-contained Streamlit application for audio transcription, speaker detection, sentiment analysis, topic detection, and Q&A using AssemblyAI. The code is well-structured with reusable display functions (display_transcription, display_summary, display_speakers, etc.) that can be extracted as a template for any audio analysis dashboard. Demonstrates clean separation of API calls and UI rendering.
A modular RAG generator that produces podcast scripts with citations from retrieved documents. It uses LlamaIndex for retrieval and a custom prompt template for script generation. The class is parameterized (model, temperature, chunk size) and could be reused in any document-to-podcast pipeline. Demonstrates a clean pattern for combining retrieval with structured output generation.
A reusable utility for creating, updating, and validating datasets on the Hugging Face Hub. Provides functions like init_dataset, add_rows, and validate_by_template that abstract away Hub API details. Can be extracted and used in any project that needs to programmatically manage HF datasets. Demonstrates clean error handling and template-based validation.
MCP tools for indexing and searching documents using Pixeltable's declarative data pipeline. The tools (index_document, search_documents) are self-contained, use Pydantic for input/output schemas, and demonstrate a clean pattern for building MCP servers that wrap a vector database. Could be adapted for any document indexing use case.