AI Agents & Automation

LeFlow is an open-source tool-flow designed to bridge the gap between TensorFlow deep neural networks and synthesizable hardware, specifically FPGAs. It achieves this by integrating Google's XLA compiler with the LegUp high-level synthesis tool, enabling the automatic generation of Verilog code from TensorFlow specifications. This facilitates the deployment of deep neural networks on FPGAs, offering a flexible approach to hardware acceleration. The tool includes a testing framework with 15 building blocks to verify installation and functionality, ensuring that generated circuits match original TensorFlow results. It also provides examples ranging from simple tests to more complex applications, making it a comprehensive solution for hardware synthesis of AI models.

mistral.rs is an open-source, high-performance framework designed for fast and flexible Large Language Model (LLM) inference. It boasts zero-configuration support for any Hugging Face model, automatically detecting architecture, quantization format, and chat template. The tool offers true multimodality, handling text, vision, video, audio input, speech generation, image generation, and embeddings within a single engine. Key features include comprehensive quantization control (ISQ, GGUF, GPTQ, AWQ, HQQ, FP8, BNB), hardware-aware tuning for optimal performance, and flexible SDKs for both Python and Rust. It also provides advanced agentic features like integrated tool calling, server-side agentic loops, web search integration, and an MCP client for external tool connections. A built-in web UI simplifies interaction, making it a versatile solution for developers building AI applications.

The Unity Machine Learning Agents Toolkit (ML-Agents) is an open-source project designed to transform games and simulations into dynamic environments for training intelligent agents. It leverages deep reinforcement learning and imitation learning, offering PyTorch-based implementations for easy integration. The toolkit supports various training scenarios, including single-agent, multi-agent cooperative, and competitive setups, using algorithms like PPO, SAC, MA-POCA, and self-play. It also facilitates learning from demonstrations with BC and GAIL algorithms. ML-Agents provides a flexible Unity SDK, allowing developers to integrate it into custom scenes and add their own training algorithms. It's ideal for controlling NPC behavior, automated game testing, and evaluating game design decisions.

CrushBank is an AI-powered platform designed to transform IT operations for Managed Service Providers (MSPs) and enterprise help desks. It leverages a managed data lakehouse to mine data, refine analytics, and power intelligent AI agents and automations. The platform enables organizations to build smarter automations using their own trusted data and machine learning, achieving over 90% accuracy in ticket classification, budgeting, and prioritization. CrushBank's autonomous AI agents continuously analyze ticket data to identify reopened tickets, uncover trends, and accelerate resolution times. It integrates with existing Professional Services Automation (PSA) systems, turning them into AI-powered engines with automations, document and conversational search, internet search, insights, and document creation. Additionally, its SmartShift and SmartArchive features facilitate seamless PSA data migration and intelligent archiving of legacy systems.

Modish AI is an innovative application that transforms home design by leveraging cutting-edge AI technologies from Astria, Dalle'3, and Midjourney. Users can snap pictures of their existing spaces and instantly generate exquisite 3D room designs in various styles like Tuscan, Industrial, or Bohemian. The app provides access to an expansive catalog of tens of thousands of decor items, from furniture to color palettes, enabling endless design possibilities. Modish AI also features photo-realistic imagery with accurate lighting and textures, an Augmented Reality (AR) adventure to experience designs virtually, and cross-platform and offline design capabilities. With Astria-powered precision, it creates detailed 3D renderings with precise measurements, making interior design both accurate and thrilling.

neptune-client is a Python client designed for the Neptune app, serving as an experiment tracker specifically for foundation model training. It enables data scientists and developers to monitor, log, and manage their machine learning experiments effectively. The tool supports various ML frameworks including TensorFlow, Keras, PyTorch, XGBoost, LightGBM, and Optuna, making it versatile for different project needs. It offers features for experiment versioning, comparison, and visualization, which are crucial for iterating on models and understanding performance. This client is essential for MLOps workflows, providing a centralized system for tracking metrics, parameters, and artifacts.

TensorFlow Recommenders Addons (TFRA) is an open-source collection of projects designed to enhance TensorFlow's capabilities for building large-scale recommendation systems. It primarily introduces Dynamic Embedding Technology, which allows for trainable key-value data structures within TensorFlow, leading to better recommendation effects compared to static embedding mechanisms by avoiding hash conflicts. TFRA is compatible with native TensorFlow optimizers, initializers, CheckPoint, and SavedModel formats. It fully supports training and inference of recommender models on GPUs, including integration with TF Serving and Triton Inference Server. The project also offers support for various Key-Value implementations as dynamic embedding storage, such as cuckoohash_map and HierarchicalKV, and supports both half-synchronous and asynchronous training methods.

Pipeline AI is an AI-powered Revenue Intelligence Platform designed to provide insights and outreach capabilities for revenue teams. It securely connects to your CRM and other GTM systems to continuously ingest revenue data, including accounts, contacts, opportunities, and signals. The platform applies intelligence to calculate signals, clean data, enrich records, and add context, delivering actionable insights directly into CRM, outreach tools, and dashboards. Key features include weekly forecast calls with clear signals, visibility into pipeline and deal changes, accurate revenue planning based on live engagement, and automated CRM data cleaning and enrichment. Pipeline AI helps CROs and RevOps teams understand revenue performance, identify emerging risks, and make data-driven decisions.

Parrots is an open-source toolkit designed for Automatic Speech Recognition (ASR) and Text-To-Speech (TTS) functionalities. It supports multiple languages, including Chinese, English, and Japanese, and provides multi-speaker voice synthesis with high accuracy. Key features include a Chinese ASR model based on distilwhisper, and TTS models like GPT-SoVITS and IndexTTS2. IndexTTS2 is particularly notable for its advanced capabilities, offering zero-shot speech synthesis with emotional expression and duration control, independent control over timbre and emotion, and support for various emotion control methods including audio reference, emotion vectors, and text descriptions. The tool also supports streaming TTS for low-latency real-time audio output and command-line interface (CLI) for both ASR and TTS tasks, making it suitable for developers and researchers.

Vertex serves as a bridge to connect to Google's Gemini AI model via an OpenAI environment. This application enables users to send their OpenAI requests and receive responses that are processed by Gemini, effectively integrating Gemini's advanced capabilities into existing OpenAI workflows. It is designed for users who wish to utilize Gemini's power without leaving their familiar OpenAI setup, offering a convenient way to access a different large language model. The tool appears to be hosted on Hugging Face Spaces, suggesting an accessible and potentially community-driven platform for AI experimentation and integration.

TabAI is an AI-powered workspace designed to enhance productivity by structuring tasks, managing browser tabs, and blocking distractions. It automatically collects tasks from popular tools like Notion, Gmail, Slack, Google Calendar, and Todoist, consolidating them into a single view. The tool uses AI to understand your work context, blocking distractions and organizing tabs intelligently. Key features include automatic task collection, context-aware blocking, smart tab organization, the ability to turn any text into a task, and built-in focus sessions with a Pomodoro timer. TabAI also provides analytics to track focus time and productivity patterns, helping users develop self-awareness and improve their workflow. It's built for performance, ensuring minimal impact on browser speed, and prioritizes user privacy by processing all data locally.

CAMEL AI is an open-source community and framework dedicated to advancing multi-agent AI systems. It focuses on discovering the scaling laws of agents for various applications, including data generation, world simulation, and task automation. The platform supports building complex multi-agent systems with features like evolvability, scalability for millions of agents, statefulness with dynamic memory management, and 'code-as-prompt' design. CAMEL AI provides a comprehensive toolkit for messaging, planning, evaluation, and observability, along with a rich ecosystem of agent types, models, tools, data loaders, environments, interpreters, retrievers, and verifiers. It is built on foundational research such as CAMEL, OWL, and OASIS, making it a robust platform for researchers and developers in the AI agent space.

PlanMate is an AI-powered travel planner that revolutionizes local exploration by intelligently clustering recommendations into geographically optimized, walkable routes. This tool is ideal for crafting engaging food tours, lively bar crawls, romantic date nights, or simply discovering hidden gems in your vicinity. PlanMate's core functionality includes smart clustering of venues within 400-meter walking routes, ensuring a seamless and enjoyable experience. It also provides rich venue intelligence, enhanced by Instagram Reels integration, offering dynamic visual insights into each location. Whether you're a local looking to rediscover your city or a traveler seeking an authentic experience, PlanMate simplifies itinerary creation and maximizes your exploration time.

InventPay is a crypto payment gateway designed for websites, Telegram bots, and applications, allowing merchants to accept Bitcoin, Ethereum, USDT (ERC-20, BEP-20), and Litecoin payments. It features low transaction fees, instant settlement, and digital storefront capabilities for selling products. The platform offers API/SDK integration for developers, including JavaScript and Python SDKs, and an MCP server for AI agents. InventPay simplifies the process of integrating crypto payments with no coding required for basic setups, and provides real-time payment tracking, automatic digital product delivery, and webhook notifications for seamless operation.

TrainrAI is an AI-powered fitness coach that leverages Apple Health metrics to deliver personalized, data-driven workout plans. It generates day-by-day strength and cardio sessions, customized to individual recovery, workload, and goals. The tool offers in-depth analysis using metrics like VO2 Max trends, HR zones, and running form signals, ensuring science-backed adaptation with features like Training Capacity and ACWR safety bands. TrainrAI prioritizes privacy with on-device processing and no backend storage of health data. It provides evolving goals, wellness scorecards, and context-aware chat for substitutions and modifications, all accessible via a native Apple Watch experience.

FOUND is an AI sourcing engine designed to drastically reduce the time recruiters spend on candidate sourcing. It transforms the traditional 14-hour process into just 15 minutes by leveraging advanced LLMs to understand voice or text briefs. Recruiters can simply describe a role or upload a job description, and FOUND instantly maps relevant companies and talent pools. The tool allows for iterative refinement, learning from user feedback to sharpen shortlists. It provides ATS-ready CSV exports and can handle personalized outreach, expanding talent reach beyond platforms like LinkedIn. FOUND aims to save recruiters significant time and budget by delivering context-aware, ranked shortlists without manual filtering or keyword bias.

MyCody is a personal career companion designed to help individuals navigate their professional journey. It serves as a journaling app to centralize career management, allowing users to effortlessly log daily achievements, goals, and learning experiences. Beyond simple logging, MyCody leverages advanced AI to analyze reflections, providing personalized and actionable feedback to help users understand their career progress, ace performance reviews, and plan future career moves. The platform also features goal-setting capabilities to align daily activities with long-term aspirations, customizable tags for organizing reflections by skill or project, and mobile-friendly access. It aims to foster self-awareness and leadership skills by encouraging consistent reflection and providing AI-powered insights.

Oz by Warp is an orchestration platform designed for cloud agents, enabling developers to build, deploy, and manage AI agents at scale. It offers a programmable agent stack that supports various triggers like cron jobs, webhooks, and APIs, making it versatile for different automation needs. The platform ensures agents are auto-tracked, auditable, and collaborative, providing a unified control plane for best-in-class steerability. Users can run agents on their own infrastructure or Warp's, and the platform supports multi-repo changes and integration with leading AI models like Claude, Codex, and Gemini. Oz provides tools like a CLI, SDK, and API for building custom agent automations and offers both quick-start options for web app deployment and build-your-own solutions for local or cloud agents.

ChatGPT-Discord-Bot is an open-source project designed to seamlessly integrate ChatGPT and DALL·E 2 functionalities directly into Discord servers. This tool aims to boost team collaboration, communication, and efficiency by allowing users to interact with AI models within their familiar Discord environment. Users can initiate conversations with ChatGPT using the `/chat` command and generate images with DALL·E 2 via the `/imagine` command. The bot also includes a `/reset` command to clear conversation history. The project provides detailed installation steps, including obtaining OpenAI and Discord API tokens, setting up the bot on Discord, and deploying it on platforms like Replit with environment variables and a CronJob for continuous operation.

chatgpt-plugin is a robust plugin designed for Miao-Yunzai / Yunzai-Bot, leveraging the Chaite core to provide advanced intelligent chat capabilities. It supports multiple model channels, tools, processors, triggers, RAG (Retrieval Augmented Generation), and a management panel. The plugin facilitates high-concurrency conversations, knowledge base retrieval, simulated human chat, and memory management locally, with an option to connect to Chaite Cloud for online channels and tools. Key features include multi-channel support with load balancing, advanced message adaptation for various input types, group context management, and a pseudo-human mode for more natural interactions. It also offers both visual and command-line control, along with automatic updates.

chatglm-openai-api is an open-source project that offers an OpenAI-compatible API for various large language models, specifically ChatGLM-6B, ChatGLM2-6B, and Chinese Embeddings Models. This tool simplifies the integration of these powerful models into existing applications by providing a standardized API interface, similar to what developers are accustomed to with OpenAI. It supports loading models from Hugging Face and running inference on GPUs, with options for local loading and multi-GPU inference. The project also includes advanced features like ngrok and Cloudflare tunnel integration for exposing the API, making it accessible for development and deployment. It's designed for developers looking to leverage these specific models with ease.

document.ai offers a universal local knowledge base solution built upon vector databases and GPT-3.5. It enables users to transform their local answer datasets into vector embeddings, which are then stored in a vector database. When a user inputs a query, the system converts the question into a vector, retrieves the most similar answers from the database, and then uses GPT-3.5 to optimize the overall structure and coherence of the response. This approach is particularly beneficial for chat-based scenarios like customer service, where a more natural and contextually relevant reply is desired. The project also explores advanced techniques for data optimization, such as addressing inaccurate query data, leveraging self-trained embedding models, and considering fine-tuning GPT models for specialized domains.

FarmWise, acquired by Taylor Farms, specializes in advanced agricultural technology, leveraging AI and computer vision for precision weeding and cultivation. Their flagship product, Vulcan, is a next-generation intra-row weeder designed for durability and precision. Vulcan utilizes FarmWise’s industry-leading plant image capture, proprietary machine learning, and powerful processing capabilities to deliver optimal weed removal in various field conditions, day or night, and across different crop growth stages. The machine features a modular architecture for easy maintenance, self-calibrates, and receives regular over-the-air software updates. Growers can fine-tune weeding precision from the cab, enabling maximum uptime and productivity, with capabilities of covering up to 3 acres per hour.

ChatGPT-wechat-bot is an open-source project that enables users to create a WeChat bot powered by ChatGPT. This tool allows for continuous, context-aware conversations, meaning the bot remembers previous interactions. Users can reset the conversation context using a predefined keyword, ensuring fresh starts when needed. The bot integrates seamlessly into WeChat group chats, responding when mentioned, and can also be configured to respond to specific keywords in private chats. It requires an OpenAI API key and can be set up with a reverse proxy for API calls. The project provides clear instructions for installation and configuration, including Docker deployment options for the API proxy.