AI Agents & Automation

Trankit is a light-weight, transformer-based Python toolkit designed for multilingual Natural Language Processing (NLP). It offers a trainable pipeline for fundamental NLP tasks across more than 100 languages, and includes 90 downloadable pretrained pipelines for 56 languages. Trankit outperforms other state-of-the-art multilingual toolkits like Stanza in various tasks, including sentence segmentation and dependency parsing, while maintaining efficiency in memory usage and speed. Key features include an Auto Mode for automatic language detection, a command-line interface for ease of use, and support for tasks such as tokenization, part-of-speech tagging, morphological feature tagging, dependency parsing, and named entity recognition. It also allows users to build and share customized pipelines.

WideLabs specializes in delivering sovereign AI infrastructure tailored for businesses. The platform provides robust GPU cloud services, enabling companies to run demanding AI workloads efficiently. Beyond infrastructure, WideLabs also develops and integrates proprietary AI models, offering advanced capabilities for various business needs. Their end-to-end solutions ensure comprehensive support from deployment to ongoing management, addressing complex challenges in generative AI, computer vision, and predictive algorithms. WideLabs aims to create a significant impact on individuals, institutions, and companies by leveraging cutting-edge AI technologies.

Unsloth is an open-source platform designed for training and running a wide array of open models, including Gemma 4, Qwen3.5, DeepSeek, and gpt-oss, directly on local machines. It offers a user-friendly web UI, Unsloth Studio, for easy interaction, alongside a code-based version, Unsloth Core. The tool boasts significant performance improvements, enabling up to 2x faster training with up to 70% less VRAM, without compromising accuracy. It supports various model types including text, audio, embedding, and vision models, and provides features like model inference, export, tool calling, and code execution. Unsloth also includes advanced training capabilities such as reinforcement learning, custom Triton kernels, and data recipes for dataset creation from diverse file types.

UltraRAG is a lightweight RAG development framework based on the Model Context Protocol (MCP) architecture, designed for both research exploration and industrial prototyping. It standardizes core RAG components like Retriever and Generation as independent MCP Servers, allowing for precise orchestration of complex control structures such as conditional branches and loops through simple YAML configuration. The platform features a visual RAG Integrated Development Environment (IDE) with a Pipeline Builder that supports bidirectional real-time synchronization between canvas construction and code editing. This enables granular online adjustments of pipeline parameters and prompts, along with an Intelligent AI Assistant for structural design, parameter tuning, and prompt generation. UltraRAG aims to lower the barrier to entry for building RAG systems and accelerate deployment, offering one-click conversion of logic flows into interactive dialogue systems and integrated knowledge base management.

OpenManus-RL is an open-source initiative, collaboratively led by Ulab-UIUC and MetaGPT, dedicated to advancing reinforcement learning (RL) tuning for large language model (LLM) agents. Inspired by successful RL tuning in models like Deepseek-R1, this project explores novel algorithmic structures, diverse reasoning paradigms, and sophisticated reward strategies. It supports rigorous testing on agent benchmarks such as GAIA, AgentBench, WebShop, and OSWorld, with all progress and tuned models openly shared. The platform integrates advanced RL algorithms like PPO and DPO through the Verl submodule, offering efficient and flexible training capabilities. It also provides a simplified library for Supervised Fine-Tuning (SFT) and GRPO tuning, making it a comprehensive solution for researchers and developers looking to push the boundaries of agent reasoning and tool integration.

ai-agents-masterclass is a comprehensive GitHub repository designed to accompany an AI Agents Masterclass video series. It offers all the code and resources used in the YouTube series, enabling developers to follow along and build their own AI agents. The masterclass focuses on empowering developers to leverage AI agents for transforming businesses and creating sophisticated software. The repository includes examples for building agents with LangChain, LangGraph, n8n, and other technologies, covering topics from basic agent creation to RAG agents, task management, and deployment. It serves as a practical guide for anyone looking to dive deep into AI agent development.

AdderNet is an innovative AI framework designed to significantly reduce computation costs in deep neural networks, particularly convolutional neural networks (CNNs), by replacing traditional multiplications with more efficient additions. This is achieved by using the L1-norm distance between filters and input features as the output response. The framework demonstrates impressive performance, achieving 74.9% Top-1 accuracy and 91.7% Top-5 accuracy using ResNet-50 on the ImageNet dataset, all without any multiplication operations in the convolution layer. It also shows strong classification results on CIFAR-10 and CIFAR-100 datasets, as well as competitive super-resolution and adversarial robustness benchmarks. The project provides code for training and evaluation on these datasets, making it a valuable resource for researchers and developers focused on efficient deep learning.

adk-go is an open-source, code-first Go toolkit designed for building, evaluating, and deploying sophisticated AI agents with flexibility and control. It provides a modular framework that applies software development principles to AI agent creation, simplifying the orchestration of agent workflows from simple tasks to complex systems. While optimized for Gemini, ADK is model-agnostic and deployment-agnostic, ensuring compatibility with various frameworks. This Go version is particularly suited for developers creating cloud-native agent applications, capitalizing on Go's inherent strengths in concurrency and performance. Key features include idiomatic Go design, a rich tool ecosystem for diverse agent capabilities, and strong support for containerization and deployment in environments like Google Cloud Run.

AgentGym-RL is a comprehensive framework designed for training Large Language Model (LLM) agents to excel in long-horizon, multi-turn interactive decision-making tasks using reinforcement learning. It addresses challenges in existing methods by offering a modular system that supports a wide array of real-world scenarios and integrates mainstream RL algorithms. The framework introduces ScalingInter-RL, a progressive horizon-scaling strategy that balances exploration and exploitation, leading to stable and efficient optimization. It includes diverse environments like Web Navigation, Deep Search, Digital Games, Embodied Tasks, and Scientific Tasks, and supports various training paradigms beyond online RL, such as SFT, DPO, and AgentEvol. AgentGym-RL also provides a visualized interactive user interface for analyzing interaction trajectories.

AnglE is an open-source library designed for training and inferring state-of-the-art BERT/LLM-based sentence embeddings. It utilizes an angle-optimized approach, offering various loss functions like AnglE loss, Contrastive loss, CoSENT loss, and Espresso loss. The library supports both BERT-based and LLM-based models, including bi-directional LLMs, and facilitates single-GPU and multi-GPU training. AnglE has achieved SOTA performance on benchmarks like STS and MTEB, with models trained using AnglE reaching top positions. It provides a flexible framework for researchers and developers to build and deploy high-quality sentence embedding models.

The University of Waterloo's Data & Artificial Intelligence Institute (Waterloo.AI) is a multidisciplinary research institute dedicated to advancing AI for economic prosperity and quality of life. It focuses on developing intelligent systems for various applications, including disease detection, language understanding, and vehicle navigation. The institute actively collaborates with industry partners to bridge the gap between academic research and practical, deployable AI solutions. Waterloo.AI aims to foster innovation and talent in the AI field, contributing to real-world impact through its research and partnerships.

Wayve is at the forefront of autonomous driving technology, developing a general-purpose driving intelligence that leverages embodied AI. This innovative software learns from real-world data, enabling it to scale across diverse vehicle types, geographical locations, and applications. The Wayve AI Driver is a mapless, vehicle-agnostic solution designed to unlock all levels of driving automation. It focuses on unparalleled safety, adapting to unexpected situations, and offers universal compatibility with various sensors and hardware. Wayve collaborates with leading automakers and technology pioneers to deliver reliable, real-world autonomy that meets high standards for safety, scale, and innovation.

Anon provides a comprehensive benchmark for assessing a website's readiness for AI agents. It scans your domain to evaluate key areas such as signup flow, robots.txt configuration, API documentation, and LLM visibility, generating a score out of 100. This score helps identify gaps before they impact AI-driven customer acquisition. The platform offers detailed breakdowns and competitive comparisons, highlighting critical areas like programmatic agent onboarding paths, agent discovery files (e.g., /.well-known/agent.json), and the visibility of pricing information within API documentation. Anon emphasizes that agent readiness is crucial for capturing AI-driven signups and revenue in the evolving agent economy.

AlpacaEval is an automatic evaluator designed for instruction-following language models, providing a fast, cheap, and highly correlated alternative to human evaluation. It boasts a Spearman correlation of 0.98 with ChatBot Arena, costing less than $10 of OpenAI credits and running in under 3 minutes. The tool offers precomputed leaderboards for common models, an automatic evaluator validated against 20K human annotations, and a toolkit for building advanced automatic evaluators with features like caching, batching, and multi-annotators. It also includes 20K human evaluation data and a simplified AlpacaFarm evaluation dataset. AlpacaEval is particularly useful for rapid model development and iterative testing, though it cautions against replacing human evaluation for high-stakes decision-making due to potential biases and limitations in instruction representativeness.

autogen-ui offers a web-based user interface for AutoGen, a powerful framework designed for building multi-agent LLM applications. This tool provides a simple chat interface that allows users to interact with predefined agent teams, streamlining the process of developing and testing AI-driven workflows. The UI is built using Next.js, with web APIs powered by FastAPI, ensuring a responsive and efficient experience. It includes a manager for running tasks and streaming results to the client. While a starting point, it demonstrates how to build interfaces using the AutoGen AgentChat API and serves as a foundational example for more complex multi-agent system development.

Auto-Deep-Research is an open-source, fully-automated personal AI assistant designed as a cost-effective alternative to OpenAI's Deep Research. Built on the AutoAgent framework, it boasts high performance on the GAIA Benchmark and offers universal LLM support, seamlessly integrating with a wide range of models including OpenAI, Anthropic, Deepseek, vLLM, Grok, and Huggingface. The tool supports both function-calling and non-function-calling interaction LLMs and handles file uploads for enhanced data interaction. Users can get started instantly with a simple command, requiring zero configuration for an out-of-the-box experience. It aims to provide a personal assistant at a much lower cost, leveraging pay-as-you-go LLM API keys.

AnyTool is a universal tool-use layer designed to enhance AI agents' interaction with various tools. It addresses critical challenges in agent automation, such as overwhelming tool contexts, unreliable community tools, and limited capability coverage. AnyTool offers lightning-fast tool retrieval through smart context management and zero-waste processing, ensuring tools are instantly ready. Its self-evolving orchestration adapts to tool ecosystems, maintaining performance from 10 to 10,000 tools. The platform also provides universal tool automation with quality-aware selection, reliability tracking, and safety controls. It supports a multi-backend architecture, extending capabilities beyond web APIs to include system operations, GUI automation, and deep research, making it easy to integrate with any AI agent.

Aphrodite Engine is an inference engine designed to optimize the serving of HuggingFace-compatible large language models (LLMs) at scale. Leveraging vLLM's Paged Attention technology, it provides high-performance model inference for multiple concurrent users. Developed through a collaboration between PygmalionAI and Ruliad, Aphrodite serves as the backend engine powering their chat platforms and API infrastructure. Key features include continuous batching, efficient K/V management, optimized CUDA kernels, and extensive quantization support (AQLM, AWQ, GPTQ, etc.). It also offers distributed inference, 8-bit KV Cache, modern sampler support, speculative decoding, and multimodal capabilities. The engine supports Linux and Windows (WSL2) with Python 3.9 to 3.12, and requires CUDA >= 12, supporting a wide range of GPUs including AMD, Intel, Google TPU, and AWS Inferentia.

ASearcher is an open-source framework designed for large-scale online reinforcement learning (RL) training of search agents, aiming to advance Search Intelligence to expert-level performance. It provides model weights, detailed training methodologies, and data synthesis pipelines, making it fully committed to open-source development. Key features include a prompt-based LLM agent for autonomous QA pair generation, a fully asynchronous agentic RL framework that decouples trajectory collection from model training, and the ability to enable long-horizon search with tool calls exceeding 100 rounds. ASearcher achieves cutting-edge performance on challenging QA benchmarks like GAIA, xBench-DeepSearch, and Frames, demonstrating substantial improvements through RL training. It also offers comprehensive guidance for building and training customized agents.

attorch offers a collection of PyTorch's neural network modules, re-implemented in Python using OpenAI's Triton. The project's core goal is to provide an easily hackable, self-contained, and readable set of deep learning operations, maintaining or improving efficiency compared to standard PyTorch implementations. It serves as an accessible starting point for developers looking to create custom deep learning operations without the speed limitations of pure PyTorch or the complexity of writing CUDA kernels. Unlike many Triton-powered frameworks focused on Transformers, attorch includes layers for diverse applications like computer vision. It supports both forward and backward passes, making it suitable for training and inference, and offers an interface with PyTorch fallback for seamless integration.

awesome-agents is a comprehensive, curated list of open-source tools and products designed for building AI agents. This resource is invaluable for developers and researchers looking to explore and implement AI agent technology. It categorizes tools into various sections, including Frameworks, Testing and Evaluation, Software Development, Research, Conversational/General Agents, Game/Simulation, Knowledge Management, Automation, Browser, and Multimodal. The list features prominent frameworks like LangChain, AutoGen, and CrewAI, alongside specialized tools for testing, code generation, and research. It serves as a central hub for discovering cutting-edge solutions and fostering collaboration within the AI agent development community.

Brilliant Labs is dedicated to fostering an open-source ecosystem, providing resources and tools for developers and creatives to innovate and shape the future. Their flagship product, Halo, is an open-source glasses platform designed for curious and creative individuals. Halo features a color microOLED display, bone conduction speakers, and an ultra low-power Alif B1 processor with a NPU for on-device AI. It includes an optical sensor for AI inference, microphones with audio activity detection, and a 6-axis IMU. Running on ZephyrOS with a Lua interface, Halo offers cross-platform mobile app connectivity and a cloud-based AI agent named Noa, which handles real-time, multimodal conversations and remembers past interactions to personalize experiences.

BitBLAS is an open-source library designed to facilitate efficient mixed-precision DNN model deployment on GPUs. It specializes in mixed-precision BLAS operations, particularly for $W_{wdtype}A_{adtype}$ quantization in large language models (LLMs). Key features include high-performance matrix multiplication for both GEMV and GEMM, supporting various mixed-precision types like FP16xFP8/FP4/INT4/2/1 and INT8xINT4/2/1. BitBLAS also offers auto-tensorization for TensorCore-like hardware instructions and provides integrations with popular frameworks such as PyTorch, GPTQModel, AutoGPTQ, vLLM, and BitNet-b1.58. Based on techniques from the "Ladder" paper, it allows for customizing mixed-precision DNN operations via a flexible DSL (TIR Script).

bitsandbytes is a powerful library designed to make large language models (LLMs) more accessible through k-bit quantization for PyTorch. It significantly reduces memory consumption during both inference and training, allowing for more efficient use of computational resources. The library provides three core features: 8-bit optimizers that use block-wise quantization to maintain 32-bit performance with reduced memory, LLM.int8() for 8-bit quantization enabling large language model inference with half the memory and no performance degradation, and QLoRA for 4-bit quantization, which facilitates LLM training with memory-saving techniques without compromising performance. It includes quantization primitives for 8-bit and 4-bit operations, along with 8-bit optimizers, making it an essential tool for developers working with large-scale AI models.