AI Agents & Automation

LlamaFarm is an open-source AI platform designed for deploying AI models, agents, databases, RAG, and pipelines locally or remotely. It emphasizes complete privacy, ensuring data never leaves the user's device, and eliminates API costs by utilizing open-source models. The platform is offline-capable once models are downloaded and is hardware-optimized for GPU/NPU acceleration on Apple Silicon, NVIDIA, and AMD. Users can build RAG applications, train custom classifiers, detect anomalies, and perform document processing. It offers a desktop app for instant setup, a CLI for development, and a Designer web interface for project management, RAG configuration, and prompt engineering.

llm-applications offers a comprehensive guide and resources for building Retrieval Augmented Generation (RAG) based Large Language Model (LLM) applications ready for production. This open-source project, hosted on GitHub, details how to develop such applications from scratch, scale their core components like loading, chunking, embedding, and serving, and evaluate different configurations for optimal performance. It also covers implementing LLM hybrid routing to integrate both open-source and closed LLMs, and serving applications in a scalable and highly available manner. The guide includes practical setup instructions for API keys (OpenAI, Anyscale Endpoints), environment configuration, and data management, making it a valuable resource for developers looking to productionize their AI solutions.

LLM4Rec-Awesome-Papers is a meticulously curated collection of academic papers and resources focused on the intersection of large language models (LLMs) and recommender systems. This GitHub repository serves as an invaluable resource for researchers, data scientists, and developers working in the field of AI and recommendation technology. It categorizes papers based on whether they involve 'No Tuning' or 'Supervised Fine-Tuning' of LLMs, offering insights into various approaches to integrating LLMs into recommendation systems. The list is continuously updated, ensuring users have access to the latest advancements and research findings. It also includes related surveys and common datasets, making it a comprehensive hub for staying current with the rapidly evolving landscape of LLM-enhanced recommendation systems.

The llm_distillation_playbook offers a comprehensive guide to best practices for distilling large language models (LLMs) into smaller, more efficient counterparts suitable for production applications. It targets engineers and ML practitioners with deep learning fundamentals and LLM familiarity. The playbook covers key concepts like teacher and student models, and provides practical advice across 12 best practices, including understanding smaller model limitations, building robust logging infrastructure, defining clear evaluation criteria, and maximizing teacher model quality. It also emphasizes the importance of diverse datasets, starting simple, and continuous monitoring in production environments. The document draws from experiences at Google and Predibase, aiming to systemize recommendations for effective LLM refinement.

Magpie is an open-source project designed for alignment data synthesis, enabling the generation of high-quality synthetic data for training large language models (LLMs). Unlike traditional methods that depend on prompt engineering or seed questions, Magpie leverages the auto-regressive nature of aligned LLMs to generate both user queries and corresponding LLM responses from scratch. This innovative approach allows for scalable data creation by prompting LLMs with their pre-query templates. The tool supports various LLM families, including Llama, Qwen, Phi, and Gemma, and offers scripts for batched SFT data generation, multi-turn conversation extension, and comprehensive dataset filtering and tagging. Magpie aims to democratize AI by making high-quality alignment data accessible and transparent.

local-rag is an open-source tool designed for Retrieval Augmented Generation (RAG) using open-source Large Language Models (LLMs). It allows users to ingest various file types, including local files, GitHub repositories, and websites, all without relying on third-party services or sending sensitive data outside their network. A key differentiator is its support for offline embeddings and LLMs, eliminating the need for external APIs like OpenAI. The tool features streaming responses, conversational memory, and chat export capabilities, making it suitable for secure, local RAG implementations.

medAlpaca is an open-source project offering large language models (LLMs) meticulously fine-tuned for medical question-answering and dialogue applications. Expanding on the foundations of Stanford Alpaca and AlpacaLoRA, its core objective is to deliver a diverse array of open-source language models, thereby facilitating the seamless development of medical chatbot solutions. These models are trained using an extensive collection of medical texts, including flashcards, wikis, and dialogue datasets, compiled into the 'Medical Meadow' dataset. The project provides detailed instructions for getting started, including environment setup and training procedures, along with benchmarks on USMLE self-assessments. It is intended for research purposes only and not for clinical use.

mcp_agent_mail serves as an asynchronous coordination layer for AI coding agents, exposed as an HTTP-only FastMCP server. It enables agents to register temporary-but-persistent identities, manage inboxes and outboxes, and maintain searchable message histories. A key feature is its advisory file reservation system (leases), which helps prevent agents from overwriting each other's work or encountering unexpected diffs. The system is backed by Git for human-auditable artifacts and SQLite for indexing and queries, ensuring transparency and efficient data management. It's designed for FastMCP clients and CLI tools like Claude Code, Codex, and Gemini CLI, facilitating coordinated efforts across multiple codebases and reducing the need for human liaison between different AI workstreams.

MaxKB, or Max Knowledge Brain, is an open-source platform designed for building enterprise-grade AI agents. It integrates Retrieval-Augmented Generation (RAG) pipelines, enabling features like direct document uploading, automatic online document crawling, text splitting, and vectorization to reduce large model hallucinations and enhance smart Q&A interactions. The platform boasts a powerful workflow engine, function library, and MCP tool-use capabilities for orchestrating AI processes in complex business scenarios. MaxKB is model-agnostic, supporting both private (DeepSeek, Llama, Qwen) and public (OpenAI, Claude, Gemini, MiniMax) large language models, and offers native multi-modal support for text, image, audio, and video. It facilitates zero-coding rapid integration into third-party business systems, making it ideal for intelligent customer service, corporate knowledge bases, and academic research.

MotionDirector is an open-source project designed for motion customization within text-to-video diffusion models. It allows users to adapt existing models to generate diverse videos with specific motion concepts, such as various sports activities (lifting weights, riding horses, playing golf) or cinematic camera movements (dolly zoom, zoom in/out). The tool supports customizing both appearance and motion in video generation, and can animate images with learned motions. It provides scripts for training MotionDirector on single or multiple video clips and for inference with pre-trained models, making it a versatile tool for researchers and developers in AI video generation.

Fabrice AI is a conversational AI tool designed to help users generate insights and ideas. It features a chat interface that enables interactive discussions on a wide range of topics, including investment strategies and climate change. Users can explore complex questions about marketplace dynamics, making it a valuable resource for deepening knowledge and enhancing discussions. The tool aims to provide an interactive platform for users to engage with AI to gain new perspectives and information.

Multi-Agents-Debate (MAD) is a pioneering framework designed to explore and enhance the debating capabilities of Large Language Models (LLMs). It addresses the limitations of self-reflection in LLMs, such as bias, rigidity, and lack of external feedback, by introducing a multi-agent debate interaction. The framework posits that 'truth emerges from the clash of adverse ideas,' allowing agents to correct each other's distorted thinking, complement resistance to change, and provide mutual external feedback. Experiments demonstrate that MAD brings significant and consistent improvements in challenging tasks like Counterintuitive QA and Commonsense Machine Translation, showcasing its potential to exploit more of LLMs' capabilities and mitigate issues like degeneration of thoughts.

NVIDIA NeMo is a scalable generative AI framework designed for researchers and PyTorch developers focusing on Large Language Models (LLMs), Multimodal AI, and Speech AI, including Automatic Speech Recognition (ASR) and Text-to-Speech (TTS). It provides tools to efficiently create, customize, and deploy new AI models by leveraging existing code and pre-trained model checkpoints. The framework supports various speech models and has seen recent updates in areas like streaming speech recognition, multilingual TTS, and conversational AI. NeMo is open-source and requires Python 3.12+ and PyTorch 2.6+ with an NVIDIA GPU for training.

Frequently Ai is an AI assistant specifically designed to empower Amazon businesses by offering comprehensive insights and actionable solutions. The tool focuses on improving operational efficiency and boosting sales performance within the Amazon ecosystem. It helps users to delve into critical metrics, providing a clear understanding of their business's health and identifying areas for improvement. By leveraging AI-driven analytics, Frequently Ai enables data-driven decision-making, allowing Amazon sellers to optimize their strategies and achieve better results. This tool is tailored to simplify complex data, making it accessible and useful for enhancing overall business performance on the Amazon platform.

Momenta is a leading autonomous driving technology company focused on developing AI solutions for vehicles. They employ a unique scalable path combining a data-driven approach with iterating algorithms, referred to as their "flywheel approach." Their product strategy includes both mass-production-ready highly autonomous driving solutions, such as Mpilot and MSD for L2 to L2++ assisted driving features, and scalable robo-taxi services targeting full autonomy. These solutions are designed to enhance the safety, convenience, and efficiency of mobility, adapting across different vehicle models and market segments. Momenta's offerings aim to transform urban mobility by providing seamless, reliable, and comfortable rides in complex urban environments.

Nebulai is a leading marketplace designed to connect businesses with both verified human AI experts and intelligent AI agents. The platform acts as a one-stop solution for AI talent acquisition, agent deployment, and accessing enterprise AI solutions. It facilitates the hiring of various AI professionals, including data scientists, machine learning engineers, and AI consultants, while also providing production-ready AI agents for tasks like customer service automation, sales enablement, and voice interactions. Nebulai aims to streamline AI adoption for enterprises by offering AI-powered matching, secure milestone payments, and project management tools, making it easier for companies to find and deploy the right AI talent and solutions.

smartofai is at the forefront of developing advanced AI agent solutions designed to foster seamless collaboration between humans and artificial intelligence. The company envisions a future where AI agents are integral to enhancing productivity and streamlining workflows in various work environments. Their flagship product, My Digital Colleague, is specifically engineered to revolutionize teamwork by providing intelligent AI assistance directly within teams. This tool aims to empower organizations to achieve greater efficiency and innovation by leveraging AI to support and augment human capabilities, making collaborative work more effective and productive.

Rivit is a no-code AI tool building platform designed to empower users to create custom AI tools without the need for extensive coding knowledge. The platform integrates with Large Language Models (LLMs), providing a foundation for developing sophisticated AI applications. It offers various subscription options to cater to different user needs. Rivit simplifies the process of AI tool creation, making advanced AI capabilities accessible to a broader audience. This platform is particularly useful for individuals and businesses looking to leverage AI for specific tasks or workflows without investing in complex development cycles. The tool is currently listed for sale, indicating a potential for new ownership and future development directions.

AgentCloud is an open-source platform designed for companies to build and deploy private Large Language Model (LLM) chat applications, similar to having a custom GPT builder. It empowers teams to securely interact with their data through a robust RAG (Retrieval Augmented Generation) pipeline, which natively supports embedding data from over 260 sources. The platform facilitates the creation of conversational apps, multi-agent process automation applications using `crewai`, and includes features for managing tools, teams, and user permissions. AgentCloud is built with a Python backend, a Next.js UI, and a Rust-based vector proxy, making it a comprehensive solution for developing and deploying AI agents.

ollama-playground is a GitHub repository showcasing a collection of interesting LLM projects developed using Ollama's open-source models. This resource is ideal for developers and researchers looking to explore practical applications of large language models. The repository includes diverse projects such as Retrieval-Augmented Generation (RAG) for PDFs (Chat with PDFs, Hybrid RAG, Multimodal RAG, Voice RAG), various agent tooling and protocols (Agent with Memory, MCP-Based Agent, ACP-Based Agents), and vision-based AI applications (Video Summarization, OCR, Emotion Detection, Object Detection, Image Search Engine). Each project comes with code examples, making it a valuable learning and development resource for building LLM applications.

Relay.app is an intuitive platform designed to automate tasks and create reliable, visual workflows using AI. It translates plain language instructions into actionable processes across more than 200 applications, including popular tools like Airtable, HubSpot, Gmail, Notion, and Slack. The platform is praised for its user-friendly interface, making it accessible even for non-programmers to build advanced workflows. Relay.app aims to provide proactive AI assistance that improves over time, working day and night to save countless hours and money for businesses. It's a game-changer for marketing, partnerships, and general office automation, offering a quick learning curve and robust capabilities for various use cases.

OpenCompass is an advanced LLM evaluation platform designed to guide users through the complex landscape of assessing large language models. It supports a diverse array of models, including Llama3, Mistral, InternLM2, GPT-4, and Claude, and offers compatibility with over 100 datasets for comprehensive benchmarking. The platform provides powerful algorithms and an intuitive interface to evaluate the quality and effectiveness of NLP models. Key features include support for various inference acceleration backends like LMDeploy and vLLM, flexible evaluation mechanisms such as CascadeEvaluator, and tools for LLM-as-judge and mathematical reasoning assessments. Users can install OpenCompass via pip or from source, and prepare datasets either offline or through automatic downloads from OpenCompass storage or ModelScope.

Lorka AI is an all-in-one AI platform designed to streamline workflows by integrating multiple leading AI chat models such as GPT, Claude, Gemini, Grok, Qwen, and DeepSeek into a single subscription. Users can switch between different AI engines within the same conversation without losing context, allowing for dynamic brainstorming, refinement, and verification. Beyond chat, Lorka AI offers a suite of dynamic features including AI Web Search for quick information retrieval, an AI Image Editor for visual content creation, and advanced tools like PDF chat, AI Translator, and AI Humanizer. It also supports a voice mode for hands-free interaction. This platform aims to provide significant savings and flexibility compared to subscribing to individual AI services, catering to professionals and students across various fields.

onnxruntime-genai provides generative AI extensions for ONNX Runtime, enabling users to run generative AI models efficiently on various devices. This API implements the complete generative AI loop for ONNX models, encompassing pre and post-processing, inference with ONNX Runtime, logits processing, search and sampling, KV cache management, and grammar specification for tool calling. It powers applications like Foundry Local, Windows ML, and the Visual Studio Code AI Toolkit. The tool supports a wide range of model architectures, including Llama, Mistral, Gemma, and Phi, and offers APIs for Python, C#, C/C++, and Java across multiple operating systems and hardware accelerations like CUDA, DirectML, and OpenVINO. Key features include multi-LoRA, continuous decoding, constrained decoding, and speculative decoding.