AI Agents & Automation

Foxtrot Systems, now part of Descartes, offers advanced AI software designed to automate and optimize last-mile logistics for enterprise fleets. The platform provides real-time fleet-wide visibility and communication, enabling businesses to operate their fleets autonomously. Key features include next-gen fleet digitization with driver apps or SDKs, a control tower dashboard, and powerful APIs for custom builds. It also offers a logistics data warehouse for analytics, client learning APIs for route planning, and real-time route execution automation. Foxtrot's solutions help maximize efficiency, improve service levels, and significantly reduce CO2 emissions and fuel consumption, making logistics a key differentiator for businesses.

Ginni AI is a sales platform designed to empower sales teams by integrating autonomous AI agents into their workflow. These agents provide real-time insights, take action, and help close more deals by eliminating guesswork and ensuring every sales action is covered. Ginni listens to every sales call, identifying winning skills, areas for improvement, and crucial customer insights. It offers features like a personal AI Sales Coach for real-time feedback, playbooks to guide sales decisions, and a Meeting Assistant to automate administrative tasks, allowing reps to focus on selling. The platform supports mixed-language conversations and phone-based sales calls, and integrates seamlessly with existing tools.

fastai is a comprehensive deep learning library designed to be both approachable for practitioners and deeply hackable for researchers. It offers high-level components that enable users to quickly achieve state-of-the-art results in common deep learning tasks, such as image classification, segmentation, text sentiment analysis, recommendation systems, and tabular models, often with just a few lines of code. For researchers, fastai provides low-level, composable building blocks, allowing for the development of new approaches without compromising ease of use, flexibility, or performance. The library features a carefully layered architecture, a new type dispatch system for Python, a GPU-optimized computer vision library, a refactored optimizer, a novel 2-way callback system, and a new data block API. It is built on PyTorch and supports easy migration from other PyTorch-based libraries.

Unisson offers AI agents designed to empower customer-facing teams, including Customer Success, Implementation, FDE, Sales Engineering, and GTM teams. The platform features two core agents: Runner, which executes customer tasks directly within your product, and Explorer, which maintains an up-to-date knowledge base to support Runner. This system aims to provide faster time to value for customers by eliminating delays and enables day-one effectiveness for team members without extensive product training. Unisson helps scale revenue by absorbing repetitive and complex product work, allowing teams to support more customers and focus on growth. It is built for enterprise, SOC2 compliant, and offers features like auditable actions, real-time answers, and an API to supercharge your own agents.

Saldor is developing AI-powered software specifically designed to reduce procedural errors in the Supplemental Nutrition Assistance Program (SNAP). The tool targets key areas like application, recertification, and change reporting, aiming to assist caseworkers in navigating the complex and frequently updated SNAP policies. By addressing issues such as rule churn, data reconciliation, interview drift, and handoff inefficiencies, Saldor seeks to minimize payment error rates. The initiative is particularly relevant given recent policy changes (HR 1) that impose significant financial penalties on states with high SNAP error rates. Saldor is currently in the design, discovery, and field research stage, actively collaborating with SNAP directors, QC staff, and caseworkers to build a tool that effectively integrates into existing agency workflows.

GPT OSS represents a groundbreaking shift in AI accessibility, released by OpenAI. This open-source GPT model family consists of two powerful variants, GPT-OSS-120B and GPT-OSS-20B, designed for reasoning and developer tasks. It aims to democratize access to advanced AI through powerful reasoning capabilities, developer-friendly features, and Apache 2.0 licensing. The models support tool use, function calling, chain-of-thought reasoning, and structured outputs, with API compatibility. GPT-OSS-120B requires a single 80GB GPU, while GPT-OSS-20B can run on consumer hardware with 16GB of memory, making it accessible for edge deployment and local inference. It includes comprehensive safety measures and a Mixture-of-Experts (MoE) architecture for efficiency.

HerculesAI Armenia offers an enterprise AI platform designed to automate complex workflows and eliminate manual bottlenecks within organizations. The platform utilizes AI-workers to understand document structures, generate business rules, and translate logic into programmatic code. This enables the verification and reconciliation of data against formalized logic, flagging discrepancies before financial transactions occur. HerculesAI is particularly effective in areas like order-to-cash accuracy automation, revenue integrity, leakage prevention, and compliance automation, providing audit-ready outputs for billing, settlement, and reporting systems. It helps reduce processing time, accelerate cash cycles, and improve margins by ensuring accuracy and audit defensibility in every workflow.

Falcon3 Mamba 7b Instruct Playground is a Hugging Face Space that provides an interactive environment to engage with the Falcon3 Mamba 7b Instruct model. Users can ask questions or have conversations with this language model, which is specifically trained to provide helpful responses. This playground is particularly useful for developers and researchers who are interested in exploring and testing the instruction-following capabilities of advanced AI models. While the space is currently paused, it offers a glimpse into the potential for conversational AI and model interaction.

Fume is an AI testing tool designed to automate end-to-end browser testing by acting as an autonomous QA team. Users record a Loom-style video describing what they want to test, and Fume generates and maintains Playwright tests. It extracts test cases in seconds, covering major user flows quickly. Fume offers self-healing tests with AI agent fallbacks for Playwright's fragility, ensuring tests continue even if scripts break. It provides free test runners in the cloud and allows full ownership of the generated Playwright code, which can be run locally or integrated into existing CI/CD pipelines. Fume also supports migrating existing Playwright, Selenium, or Cypress tests.

Rhippo is an AI tool designed to enhance team collaboration by offering dynamic knowledge and context injection. It integrates AI assistance into existing workflows, aiming to improve efficiency and communication within teams. The tool's core functionality appears to revolve around providing relevant information and contextual understanding to users, though specific features are not detailed. The website content suggests a focus on professional environments, potentially assisting with information retrieval and task management. While the exact mechanisms are not explicitly stated, the emphasis on "dynamic knowledge" and "context injection" points towards an intelligent system that adapts and provides relevant insights.

gradio_multimodalchatbot V0.0.1 is a Hugging Face Space that provides a versatile Gradio component for displaying various media types within chat message bubbles. This tool allows users to integrate text, images, audio, and video files directly into conversational interfaces, enhancing the richness and interactivity of chatbot interactions. It's designed to make chatbot experiences more engaging and informative by supporting diverse media formats. The application is hosted on Hugging Face Spaces, indicating its accessibility and potential for community-driven development and use.

NVIDIA FLARE (NVFlare) is a domain-agnostic, open-source, extensible Python SDK designed to help researchers and data scientists adapt existing machine learning and deep learning workflows to a federated paradigm. It provides a robust framework for building secure, privacy-preserving offerings for distributed multi-party collaboration. Key features include support for both deep learning and traditional machine learning algorithms (e.g., PyTorch, TensorFlow, scikit-learn, XGBoost), horizontal and vertical federated learning, and built-in federated learning algorithms like FedAvg and FedProx. NVFlare also supports multiple server and client-controlled training and validation workflows, data analytics, and privacy preservation techniques such as differential privacy and homomorphic encryption. It facilitates seamless transition from ML/DL to FL with minimal code changes and offers deployment options on cloud and on-premise.

HKB Tech is an AI and DeepTech company dedicated to fostering global growth through advanced AI solutions. Their core expertise lies in patented Agentic AI/ML and proprietary DeepTech intellectual property. The company focuses on delivering strategic projects and solutions for a diverse clientele, including government entities, large-scale mega-projects, and various enterprises across multiple sectors. HKB Tech aims to leverage its specialized AI capabilities to address complex challenges and drive innovation in different industries.

Invoice Matchpoint by Dodocs AI is an Accounting OS designed to automate the tedious tasks of document chasing, statement reconciliation, and month-end closing for accounting firms. The platform captures and classifies documents like invoices, receipts, and bank statements from various channels including email and WhatsApp, with an 87% auto-publish rate. Its AI extracts data with high confidence, verifies it against transactions, and auto-publishes to accounting software like QuickBooks Online. Dodocs AI also features 'Chase Loops' to automatically follow up on missing documents and 'Reconciliation & Fraud Alerts' to match invoices against bank statements, detecting duplicates and anomalies before they impact the books. This comprehensive solution aims to put accounting on autopilot, allowing teams to focus on growth.

Spydr Memory MCP is introduced as the first multimodal, interoperable context engine designed for any AI client. It addresses the challenge of fragmented information and context spread across various siloed applications, allowing users to maintain and transfer their context effortlessly. The tool enables AI to curate user context, transforming it into valuable starting points for discovery. Key features include AI-powered context curation, the ability to make context 'On the Go,' and breaking down information silos. Users can sign up by continuing with Google or creating a new account, with the service currently in BETA.

FinanceBrain is an AI-powered assistant designed to provide instant answers to finance and accounting questions. This tool offers 24/7 availability, ensuring users can get support whenever needed. New users are granted 5 free questions to experience the service. For continued and unlimited access, FinanceBrain offers both monthly and annual subscription plans, with a discount available for annual billing. It aims to streamline the process of finding information related to financial queries, making it a valuable resource for professionals and students alike.

Big Speak is an AI-powered tool designed to enhance audio experiences through advanced machine learning algorithms. It specializes in text-to-speech conversion, allowing users to transform written content into natural-sounding audio. Additionally, the tool provides audio transcription services, converting spoken words into text. A key feature of Big Speak is its voice cloning capability, enabling users to create custom voice models for personalized audio output. The tool aims to produce high-quality audio, catering to various needs from content creation to personalized communication. While specific pricing details are not available, the tool is described as offering both free and premium plans.

Nekton is an AI-powered automation platform designed to streamline daily tasks and workflows by allowing users to describe their tasks step-by-step in plain English. The AI then writes and runs the automation code in the cloud, eliminating the need for users to learn complicated tools or maintain infrastructure. Powered by OpenAI and ChatGPT, Nekton integrates with thousands of services and can automate various business and personal workflows, from getting notified about product mentions to automating news posting and creating checklists. It supports features like forms for user input, sharing workflows, scheduled runs, and easy customizations for highly-specific tasks, allowing for gradual automation by mixing manual and automated steps.

AI Music API offers a unified REST API for developers to generate production-quality music using various AI models, including Suno, Udio, Stable Audio, and MusicGen. It simplifies the integration process by providing a single endpoint, eliminating the need for individual model management and infrastructure setup. The platform supports rapid prototyping and production-scale applications, enabling multi-genre mixing, dynamic soundtracks, and unique track discovery. Key features include lightning-fast inference, multi-model access, enterprise-grade security, a global edge network for low latency, and detailed usage analytics. Users can control genre, mood, tempo, duration, and instruments, and generated music comes with full commercial licensing.

Ask My Book is an AI experiment specifically designed to provide answers and insights from the book 'The Minimalist Entrepreneur.' This tool allows users to interact with the book's content through an AI interface, making it easy to find specific information or get summaries of key concepts. It's particularly useful for readers who want to quickly grasp the essence of the book, review specific sections, or get immediate answers to their questions without having to reread entire chapters. The interactive nature of the tool aims to enhance the learning and comprehension experience for its users.

yourly is an innovative recruiting platform designed to bridge the gap between startups and top talent by leveraging trusted referrals. The platform empowers individuals, referred to as 'connectors,' to utilize their professional networks to identify suitable candidates for various startup roles. An AI system assists connectors in finding the best fits within their network. When a referred candidate is successfully hired, the connector earns a reward, typically ranging from $3,000 to $15,000. This model capitalizes on the fact that a high percentage of top talent comes from referrals, offering a mutually beneficial system for connectors to earn income and for startups to access high-quality candidates. The platform is currently in early access, working with over 30 startups.

Datatruck is an AI-native all-in-one transportation management system (TMS) designed for the trucking industry. It helps carriers, brokers, hybrids, enterprise, and private fleets manage operations, cut costs, and boost productivity. The platform leverages AI for automated workflows, including TruckGPT for extracting key logistics data, AI Agents for load negotiations and driver communication, and an AI Financial Assistant for reconciliation and financial insights. Datatruck aims to replace multiple spreadsheets with a consolidated solution, offering seamless data integration, cloud analytics, and over 100 seamless integrations with trucking and finance tools. It provides solutions for fleet and safety management, financial automation, and data analytics, empowering data-driven decisions and efficient scaling for logistics businesses.

Missing Studio operates as an Applied Autonomy Lab, dedicated to creating systems that enable businesses to run autonomously through AI. Their core mission is to empower a single human to operate at the scale of an entire company by leveraging autonomous systems. The studio emphasizes applied intelligence, ensuring every idea is proven in real-world applications, and practices relentless execution, moving fast and iterating hard to finish what they start. They are builders by default, taking end-to-end ownership of architecture, decisions, failures, and outcomes, with excellence as a baseline constraint for robust, scalable, and elegant solutions. Based in Bangalore, India, they aim to be The General Intelligence Company of India.

Reka AI is a frontier AI research company specializing in building unified multimodal foundation models. Unlike single-modality systems, Reka’s models are designed to process and reason across text, image, video, and audio simultaneously, providing enterprise-grade intelligence for complex, real-world workflows. Key offerings include Reka Vision for multimodal perception and reasoning across video and images, Reka Research for state-of-the-art foundation models, and Reka Speech for advanced audio understanding. Reka also offers a tiered family of models like Reka Core, Reka Flash, and Reka Spark, optimized for various use cases from high-stakes reasoning to on-device computing. The platform supports flexible deployment options including cloud, VPC, and on-premise environments.