AI Agents & Automation

DeepResearch, developed by Alibaba-NLP, is an open-source agentic large language model specifically engineered for long-horizon, deep information-seeking tasks. With 30.5 billion total parameters and 3.3 billion activated per token, it demonstrates state-of-the-art performance across various agentic search benchmarks. Key features include a fully automated synthetic data generation pipeline, large-scale continual pre-training on agentic data, and end-to-end reinforcement learning. It supports both ReAct and IterResearch-based 'Heavy' inference paradigms, allowing for rigorous evaluation of intrinsic abilities and maximum performance. The model is available for download and can be integrated via OpenRouter for inference without GPUs.

DeepSeek-MoE is an innovative Mixture-of-Experts (MoE) language model featuring 16.4 billion parameters. It utilizes a unique architecture with fine-grained expert segmentation and shared experts isolation, allowing it to achieve performance comparable to DeepSeek 7B and LLaMA2 7B while requiring only about 40% of the computations. Trained from scratch on 2 trillion English and Chinese tokens, DeepSeek-MoE provides both base and chat model checkpoints for research and commercial use. It can be deployed on a single GPU with 40GB of memory without quantization, and offers quick start guides for installation and inference using Huggingface's Transformers. The project also provides scripts for fine-tuning the models on downstream tasks, supporting both DeepSpeed and QLoRA configurations.

Gustavou is an AI-powered chef assistant designed to revolutionize home cooking. It offers personalized recipes tailored to individual preferences and dietary needs, along with weekly meal planning to simplify dinner decisions. Users can generate smart shopping lists directly from their saved recipes, ensuring efficient grocery trips. The tool also allows for recipe transformation, enabling users to convert dishes into healthier, vegetarian, or vegan versions. Gustavou provides professional cooking guidance, including step-by-step instructions and tips for mastering techniques like air frying, slow cooking, and BBQ. It aims to eliminate common kitchen struggles such as endless recipe scrolling, meal planning fatigue, and food waste, making cooking enjoyable and accessible for everyone.

Nectir AI is a leading AI infrastructure platform designed specifically for higher education institutions. It enables the creation of customizable, FERPA-compliant AI Assistants that integrate directly into Learning Management Systems (LMS). These AI Assistants provide 24/7 student support, personalized guidance, and help faculty streamline routine tasks. The platform is built to be secure, protecting student privacy and institutional data by using private endpoints and never training external models with uploaded content. Nectir AI supports various leading AI models and offers intuitive analytics for campus-wide AI usage monitoring. It aims to enhance learning outcomes, increase student success, and improve administrative efficiency across campuses.

EmbedAPI serves as a comprehensive AI integration platform designed to simplify the process of connecting to various AI models. It offers a unified API that allows developers to integrate leading AI models such as OpenAI, Anthropic, and Vertex AI quickly and efficiently. The platform aims to streamline AI development by providing a single point of access, reducing the complexity typically associated with managing multiple AI service providers. This enables faster deployment of AI capabilities into applications and services, making it an essential tool for developers looking to leverage diverse AI technologies without extensive setup.

Datrics AI offers an enterprise-grade AI platform specifically designed to automate complex workflows within the healthcare sector. It focuses on streamlining medical coding, healthcare claim adjudication, auditing processes, and ensuring compliance. By leveraging advanced AI technology, Datrics AI aims to improve accuracy, significantly reduce processing times, and eliminate costly errors commonly associated with manual claim processing. The platform provides measurable return on investment for organizations seeking to enhance efficiency and operational effectiveness in their healthcare claim management.

DeepSeek-671B-SFT-Guide offers an open-source solution for the full parameter fine-tuning of DeepSeek-V3/R1 671B models. Developed by the Institute of Automation of the Chinese Academy of Sciences and Beijing Wenge Technology Co. Ltd., this guide includes comprehensive code and scripts covering the entire process from training to inference. It also shares practical experiences, common pitfalls, and solutions encountered during model training and deployment. Key features include implemented modeling files for DeepSeek-V3/R1 training logic, support for full parameter fine-tuning using data parallelism (DeepSpeed ZeRO) and sequence parallelism, and detailed instructions for environment setup, data preparation, training, model weight conversion, and inference deployment. The guide is designed for technical users looking to fine-tune large language models efficiently.

ProtoScience is an open archive showcasing autonomous scientific discovery experiments. This platform demonstrates how machine learning can rediscover fundamental scientific laws directly from raw, real-world measurements, without any prior physics hints or textbook formulas. It employs a deterministic pipeline to analyze data and identify governing equations across diverse domains. The archive currently features 25 experiments spanning 15 domains, with 37 laws found, including 5/5 GR verified laws. It serves as a valuable resource for researchers and data scientists interested in the application of AI to scientific discovery, offering insights into how machines can independently derive complex scientific principles.

Promptise Foundry, formerly DeepMCPAgent, is an open-source framework designed for building full-stack agentic systems. It provides a comprehensive suite of tools for developers to create production-ready, secure, and scalable AI agents. The framework includes a powerful reasoning engine with 20 node types and 7 prebuilt patterns, an MCP Server SDK for multi-user, secure tool access, and an autonomous agent runtime with crash recovery and budget enforcement. It also features advanced prompt engineering capabilities, allowing prompts to be built like software with various block types, strategies, and guards. Promptise Foundry aims to simplify the development of complex AI agents by offering a unified framework that replaces multiple individual libraries.

Sukha-Focus for Remote Work is designed to enhance healthy productivity by providing a suite of tools for remote workers. The platform integrates focus music, customizable timers, and AI assistants to help users combat distractions and maintain concentration throughout their workday. It aims to create a stress-free work environment, promoting mindfulness and improving overall work-life balance. Sukha supports various productivity techniques, including the Pomodoro technique, and offers features like meeting notifications and task management to optimize work schedules and achieve professional goals. This tool is ideal for individuals looking to boost work efficiency and cultivate mindful working habits.

Halal Personal Advisor is an AI-powered platform designed to streamline halal compliance for businesses. It features a Halal Product Manager module that simplifies the management of halal products, raw materials, and supporting documentation through intelligent automation. Key capabilities include product and raw material tracking, smart document management, real-time alerts for renewals and missing information, and an AI Compliance Checker for instant validation. Additionally, the Halal AI Assistant acts as a knowledge engine, providing fast and accurate answers on halal certification requirements, fees, regulations, and recognition status across different countries and sectors. This all-in-one solution helps businesses stay organized, compliant, and competitive in the halal market.

Nunchaku is a high-performance inference engine specifically designed for 4-bit neural networks, implementing the SVDQuant post-training quantization technique. This technology allows for 4-bit weights and activations while maintaining visual fidelity, as detailed in the accompanying ICLR 2025 Spotlight paper. The engine achieves significant memory reduction, up to 3.6x for 12B FLUX.1-dev models, and offers substantial speedups, such as 8.7x over 16-bit models on a 16GB laptop 4090 GPU by eliminating CPU offloading. Nunchaku also supports various features like LoRA, ControlNet, asynchronous offloading, and compatibility with ComfyUI, making it a versatile tool for accelerating diffusion models and other AI applications.

embabel-agent is an open-source agent framework designed for the JVM, allowing developers to author agentic flows that combine LLM-prompted interactions with custom code and domain models. It features sophisticated planning capabilities, going beyond simple state machines to dynamically formulate and re-plan action sequences to achieve goals. The framework supports strong typing and object-oriented benefits, ensuring clean interaction between prompts and authored code. Key differentiators include superior extensibility, platform abstraction for consistent QoS, and design for effective LLM mixing, enabling the use of various models for different tasks, including local models for cost and privacy. Built on Spring, it integrates easily with existing enterprise functionality and offers robust testing capabilities for both unit and end-to-end agent flows. It supports both annotation-based and Kotlin DSL approaches for flow authoring.

Lovon AI Therapy is an AI-powered mental health therapy platform designed to provide personalized treatment and emotional support. Users can engage in voice conversations, similar to talking with a real person, to process thoughts and emotions, and receive compassionate guidance anytime. The platform is built with input from PhD psychologists and mental health professionals, ensuring evidence-informed support for issues like stress, anxiety, overthinking, and relationship problems. Key features include 24/7 voice therapy, daily check-ins, personalized insights after each session, and a private space for reflections. Lovon aims to make mental health support accessible and affordable, offering a 7-day free trial.

NBoost is a scalable, open-source platform designed to enhance the relevance of search results by deploying state-of-the-art transformer models. It acts as a neural proxy, sitting between a client and a search engine like Elasticsearch, to rerank results based on fine-tuned models. This allows for domain-specific neural search engines and can improve other ranked input tasks such as question answering. NBoost offers easy installation via Docker, PyPi, or Kubernetes, and provides benchmarks demonstrating significant search boost compared to traditional methods. It supports both PyTorch and TensorFlow dependencies, making it flexible for various deployment environments.

OpenBrowser is an autonomous web browsing framework built in TypeScript, designed to empower AI agents to interact with the web like a human. Leveraging Playwright, it supports leading AI models such as OpenAI, Anthropic, and Google, allowing agents to perform tasks described in natural language. Key capabilities include navigating, clicking, typing, scrolling, and extracting data without manual scripting. It features multi-model support via the Vercel AI SDK, an interactive REPL for debugging, and sandboxed execution with resource limits. OpenBrowser is production-ready, offering stall detection, cost tracking, session management, and replay recording, making it a robust solution for browser-based AI automation.

Resistant AI offers advanced document fraud detection software capable of identifying fake, tampered, or AI-generated documents across various formats like PDFs and image files. This tool helps organizations, particularly in finance, to keep fraudulent documents out of their workflows, enhancing processes like merchant onboarding, loan underwriting, tenant screening, and claims processing. Beyond documents, Resistant AI also provides an AI upgrade for existing transaction monitoring systems, offering over 80 off-the-shelf models designed by AI and fincrime experts. This allows for improved AML risk coverage, real-time APP fraud detection, and identification of synthetic identities and risky behaviors in BNPL scenarios, all without replacing current tech stacks. The platform boasts 3x more fraud detected, over 90% fewer manual reviews, and review speeds of less than 20 seconds per document.

OSWorld is an open-source benchmark designed to evaluate multimodal AI agents performing open-ended tasks in real computer environments. It offers a robust framework for researchers and developers to test and compare the capabilities of their AI agents. The platform supports various virtualization technologies like VMware, VirtualBox, and Docker, with ongoing support for cloud platforms such as AWS. Key features include parallel execution of experiments, detailed result logging with screenshots and video recordings, and tools for manual task examination. OSWorld aims to standardize the benchmarking process for AI agents, providing clear metrics for success rates across different domains like Office, Daily, and Professional tasks.

filter-pruning-geometric-median is an open-source implementation of the Filter Pruning via Geometric Median method for accelerating deep convolutional neural networks. Developed in PyTorch, this tool enables researchers and developers to reduce the computational cost and memory footprint of their models without significant loss in accuracy. It supports both network-level and layer-level sparsity configurations, offering flexibility in how pruning is applied. The repository provides detailed usage instructions for integration with PyTorch and NNI, along with scripts for reproducing results on datasets like ImageNet and CIFAR-10, making it a valuable resource for model compression research and application.

ext-apps is the official repository for the specification and SDK of the Model Context Protocol (MCP) Apps protocol, offering a standardized way to deliver interactive UIs from MCP servers directly within AI chatbots. This tool allows developers to create dynamic user interfaces such as charts, forms, and dashboards that render inline in compliant chat clients like Claude and ChatGPT. It extends the core MCP specification by enabling tools to declare UI resources, which the host then fetches and displays in a sandboxed iframe, facilitating bidirectional communication. The SDK supports app developers building interactive Views, host developers embedding these Views, and MCP server authors registering tools with UI metadata. It includes agent skills to scaffold new apps, migrate existing OpenAI apps, and convert web apps to hybrid web + MCP Apps.

AFI Explorer is a mobile application designed to revolutionize how Air Force personnel interact with official publications. Leveraging AI, the tool enables users to instantly chat, reference, and search Air Force Instructions (AFIs) directly from their mobile devices. This eliminates the need for manual navigation through lengthy documents, significantly boosting efficiency. The app aims to help users quickly and efficiently find the information they need, making complex military regulations more accessible and manageable. It's built to streamline the process of understanding and applying Air Force guidelines, providing a modern solution for accessing critical information on the go.

Otter is an open-source multi-modal model developed by EvolvingLMMs-Lab, built upon the OpenFlamingo architecture. It excels in instruction-following and in-context learning, trained extensively on the MIMIC-IT dataset, which comprises 2.8 million interleaved image-text/video instruction-response pairs. Otter supports various tasks, including scene comprehension, reasoning, and multi-round conversations, and can process both image and video inputs. The project also introduces OtterHD for fine-grained interpretations of high-resolution visual input and MagnifierBench for evaluating tiny object recognition. It provides training scripts, pre-trained weights, and supports integration with Hugging Face models.

Executive AI Assistant (EAIA) is an AI agent designed to function as a virtual executive assistant, automating various administrative tasks. It helps manage emails, schedule meetings, and triage communications based on user-defined preferences. Users can configure settings for email monitoring, response preferences, and scheduling preferences. EAIA supports both local deployment for testing and production deployment on the LangGraph Platform, offering continuous operation and integration with Google OAuth for email management. The tool provides scripts for easy setup, email ingestion, and cron job scheduling, making it suitable for individuals seeking to streamline their executive assistant responsibilities.

NSFWGirlfriend is a platform designed for adult entertainment, enabling users to create and interact with sexy AI Girls. It offers Gemini AI Girl Style Chatting and AI porn chat, providing a unique dating experience with AI companions. Key features include realistic NSFW conversations, diverse AI Girlfriend personalities, and language translation capabilities. The platform also supports virtual NSFW dates, dynamic NSFW chat filters, and NSFW AI-generated art. With chat session encryption and conversation logs, users can engage in private and personalized interactions, making every chat session tailored to their preferences.