Coding & Development

Rebellions provides advanced AI accelerators specifically designed for data centers, offering a comprehensive suite of hardware solutions such as AI cards, servers, and rack-scale systems. The company distinguishes itself with its proprietary full-stack software, which ensures native support for popular AI frameworks like PyTorch and vLLM. This integration allows for seamless deployment and optimization of AI workloads. Rebellions' hardware, exemplified by the REBEL-Quad, focuses on high-throughput compute, massive HBM3E bandwidth, and a scalable chiplet architecture, setting new standards for performance per watt. Their offerings, including RebelServer™ and ATOM™-Max Server, are built for efficient AI inference and optimized for real-world AI deployment, emphasizing scalability from individual servers to entire racks.

Ambient Scientific is a fabless semiconductor company pioneering next-generation AI processors for edge computing. Their flagship product, the GPX10, is an ultra-low power AI microprocessor designed for on-device AI applications, enabling significant cloud savings, near-zero latency, and enhanced data privacy. The GPX10 features a paradigm-shifting chip architecture with a custom AI Core (MX8) and Instruction Set Architecture, delivering superior AI throughput. It boasts ultra-low power consumption, allowing for years of always-on AI applications on a single coin cell battery, thanks to innovations like Analog Matrix Compute. The chip also offers a compact form factor, high compute density, and 100% programmability with software-definable AI cores. Ambient Scientific provides a complete SDK, including an AI Compiler, development libraries, and sample AI applications for rapid time-to-market.

Vicarius offers vRx, an advanced vulnerability remediation platform designed to automate and streamline security operations. It provides real-time and automated patching, patchless protection, and script-based remediation across applications, operating systems, and third-party software. Key features include AI-generated scripts by vuln_GPT, continuous exposure assessment, and a scripting engine for complex vulnerabilities like Log4j. The platform unifies asset and software discovery, vulnerability assessment, risk prioritization, compliance, reporting, and remediation, covering over 10,000 third-party applications. It aims to reduce manual effort, save time, and enhance system security and efficiency for IT and security teams.

Build & Ship is a deployment platform designed to simplify the process of getting applications live. It allows users to deploy any application in as little as 30 seconds with a single command, running on their own hardware. The platform supports over 15 frameworks, including Next.js, Django, Go, Rust, Flask, and Express, with automatic framework detection and Dockerfile generation. Key features include blue-green deployments for zero downtime, integrated CI/CD with GitHub, AES-256 encrypted secrets, and a self-healing node swarm architecture. It offers multiple deployment methods via CLI, macOS Finder integration, and a VS Code extension, along with a user-friendly UI for those who prefer not to use the terminal. Build & Ship aims to provide a powerful, free, and vendor-lock-in-free alternative to traditional cloud deployment services.

Deeploop Technologies specializes in developing elegant AR, VR, IoT, and AI solutions designed to feel like magic. Their offerings include AR for sales and training, virtual try-ons, visualizers, and spatial guidance. For VR, they provide simulations for practicing complex workflows in safe, measurable, and repeatable environments. In the IoT space, they build smart, connected devices that allow training on actual hardware. Additionally, their AI and Computer Vision capabilities encompass vision and analytics, pose/gesture analysis, QA, and on-device inference. They focus on delivering crisp, minimal products for various industries, bringing training, visualizations, and showrooms into immersive, realistic spaces.

Linear Logic B.V. is at the forefront of developing fully independent European drone technology, with a focus on location intelligence and autonomous operations. Their flagship product, Stormvogel, is a robust drone designed for last-mile precision with a 7 kg payload capacity and an action radius of up to 30 km. It boasts advanced features like Visual Position Hold for stable hovering, Comms Loss Recovery for autonomous path-planning, and Over-The-Air Updates for continuous optimization. The drone runs on Lilypad OS, which enables fully autonomous last-mile guidance and tactical modes without operator input. Linear Logic emphasizes local manufacturing, with all vital components, including flight controllers, ESCs, batteries, and airframes, built in the Netherlands.

Gaia is developing an active and intelligent ecosystem designed to support applications capable of learning, improving, and growing over time. While specific features are not detailed, the focus is on creating a dynamic environment for evolving AI applications. This suggests a platform aimed at fostering adaptive and self-optimizing systems, potentially for various domains. The tool aims to provide the foundational infrastructure for AI applications to develop and mature autonomously within its ecosystem.

Dataset To Model Monitor is a specialized tool hosted on Hugging Face Spaces, designed to help users keep tabs on their datasets and the models built upon them. By providing a dataset ID, users can initiate tracking and receive notifications whenever new models are trained using that specific dataset. These alerts are delivered through discussions on the Hugging Face platform, ensuring users stay informed about the evolution and application of their data. This tool is particularly useful for developers and researchers who need to monitor the impact and usage of their datasets within the AI community, facilitating better understanding and management of data-model relationships.

Clone Git Repo To Space is a utility designed to streamline the deployment of GitHub repositories to Hugging Face Spaces. Users can easily upload the contents of a Git repository by providing the Git repo URL, their Hugging Face Space ID, and a personal token. This tool is particularly useful for developers and AI practitioners who frequently work with Hugging Face's platform for hosting models and applications. It eliminates manual file transfers and complex setup procedures, offering a direct and efficient way to get code running in a Space. The integration with Hugging Face's ecosystem makes it a valuable asset for managing and sharing AI projects.

Extropic is pioneering thermodynamic computing hardware designed to be significantly more energy-efficient than traditional GPUs. Their core innovation lies in thermodynamic sampling units (TSUs) which are inherently probabilistic, making them ideal for probabilistic AI workloads. Extropic offers the XTR-0 prototype platform for developing ultra-efficient AI algorithms, featuring low-latency communication between their chips and standard processors. Additionally, they provide THRML, an open-source Python library, enabling developers to create thermodynamic algorithms and simulate their execution on TSUs. This technology aims to disrupt data centers by offering a more sustainable and powerful computing solution for the future of AI.

Github Repo To Spaces is a specialized tool designed to streamline the deployment of GitHub repositories into Hugging Face Spaces. This platform allows developers and researchers to quickly transform their codebases into interactive web applications hosted on Hugging Face. By automating the setup process, it significantly reduces the time and effort typically required for sharing AI models, demos, and educational projects. The tool is particularly useful for those looking to showcase their work in an accessible format without deep knowledge of web deployment. It supports rapid prototyping and easy sharing within the AI/ML community, making it an efficient solution for demonstrating code and models.

Github Release is an AI tool available on Hugging Face designed to automate the creation of comprehensive release notes. By analyzing commits from a specified GitHub repository branch since a given tag, it generates a formatted list of changes. Users provide their GitHub token, repository name, tag, and branch to leverage this functionality. Built with Gradio, this tool streamlines the release process for developers, ensuring that all relevant changes are captured and presented clearly, saving time and reducing manual effort in documentation.

GGUF My Lora is a specialized tool designed for developers and AI enthusiasts working with PEFT LoRA adapters. It facilitates the conversion of these adapters into the GGUF format, which is optimized for efficient inference on various hardware. Users can sign in with their Hugging Face account, specify a PEFT LoRA adapter repository, and optionally a base model. The application then swiftly converts the adapter and applies the chosen quantization, streamlining the process of preparing models for deployment or further experimentation. This tool is particularly useful for those looking to optimize their AI models for performance and compatibility within the GGML ecosystem.

DELMIA Augmented Experience, formerly Diota, is an industrial augmented reality solution designed to optimize manufacturing operations. It provides real-time digital work instructions by overlaying 3D models, work instructions, and inspection data onto real-world equipment. This technology guides operators through assembly, quality inspection, and maintenance tasks, improving accuracy, reducing errors, and standardizing work across sites. The solution offers an authoring environment to create and manage AR work instructions, leveraging existing engineering data and enabling automated updates. It supports both cloud and on-premise deployments and integrates with DELMIA Apriso and other IT systems for digital continuity.

The Otterize network mapper is a zero-config, lightweight tool designed to provide insights into Kubernetes cluster traffic without requiring extensive modifications like new CNIs or service meshes. It maps pod-to-pod traffic, pod-to-Internet traffic, and even AWS IAM traffic. Users can leverage the Otterize CLI to list traffic by client, visualize the network, export results in JSON or YAML formats, or reset remembered traffic. The tool continuously builds a communication map, which can be exported as ClientIntents Kubernetes resource files for use with the Otterize intents operator to configure network policies and Kafka ACLs.

Plandek provides engineering insights, proven frameworks, and expert guidance to accelerate the safe transition to hyper-productive AI engineering. It uses DevOps data and survey insights to reveal how work flows, where it slows down, and how to unlock real productivity. The platform helps align engineering effort to roadmap priorities, unlock developer productivity by identifying lost effort, and fast-track AI adoption and impact by detecting usage, constraints, and risks. Plandek also improves developer experience through surveys and unifies data from Jira, Git, CI/CD, and other tools to surface progress, predict risk, and reforecast delivery dates with confidence. Its AI Engineering Assistant, Dekka, analyzes data to brief teams on hidden risks, blockers, and next best actions.

Chipletti is dedicated to advancing AI compute capabilities through its high-performance AI accelerator hardware. The company specializes in developing advanced node chiplet modules designed for integration into physical machines, offering a significant boost to AI processing. This technology is engineered to enhance the efficiency and speed of AI workloads, making it suitable for demanding applications. Chipletti aims to provide robust hardware solutions that empower users to improve their AI infrastructure and achieve superior performance in their AI-driven projects.

DeviceHub is a comprehensive Connected Hardware Intelligence platform designed for companies developing and deploying IoT devices. It offers robust tools for secure, real-time IoT device management and monitoring, enabling businesses to efficiently scale their operations. The platform facilitates proactive monitoring, intelligent automation, and seamless firmware updates, including support for AOSP and Android-based devices. With DeviceHub, users can gain full control over their IoT ecosystems, minimize downtime, and leverage AI-driven analytics for deeper insights into product performance and operational efficiency. It's built to simplify IoT operations from deployment to ongoing management.

Thalo is an event sourcing runtime designed to facilitate the creation of robust, event-driven applications. It integrates WebAssembly (Wasm) through `wasmtime` for high-performance, isolated aggregate computation and utilizes `sled` as an embedded event store for fast and reliable event persistence. Key features include actor-based aggregates capable of handling over 10,000 commands/sec, projections with at-least-once guarantees, and LRU caching for optimized memory usage and fast aggregate rebuilding. Thalo supports multi-language development by compiling aggregates to Wasm, with Rust currently supported, and offers event relaying to external services like Redis streams with at-least-once guarantees. While currently unmaintained, its successor projects, SierraDB and kameo_es, continue its principles.

Tonbo is an embedded database specifically designed for serverless and edge runtimes, bridging the gap between stateless compute and persistent data. It stores data as Parquet files on S3, with coordination handled through a manifest, ensuring compute remains fully stateless. The entire storage and query engine is async-first, making it ideal for modern cloud environments. Tonbo supports rich data types and declarative schemas, allowing users to query with zero-copy RecordBatch. It runs anywhere, including Tokio, WASM, and various edge runtimes, and can serve as a storage engine for custom data infrastructure. With open formats, its standard Parquet files are readable by any tool, offering flexibility and interoperability.

Pump is an intelligent cloud platform designed to help teams monitor, optimize, and manage their cloud infrastructure across AWS, GCP, and Azure. It specializes in automating cloud cost savings, providing enterprise-level pricing, and offering comprehensive visibility into cloud spend. The platform includes features like Pump Save for automatic bill reduction, Pump View for detailed cost breakdown and forecasting, and Pump Secure for scanning against compliance frameworks and 24/7 monitoring. Pump also offers real-time incident response and actionable security insights, aiming to simplify cloud operations and reduce context-switching during outages. The platform is free for basic use, with revenue generated from cloud providers, making it a cost-effective solution for cloud management.

popmon is an open-source Python package designed to monitor the stability and quality of datasets, specifically Pandas and Spark dataframes. It generates histograms of features binned in time-slices and compares their stability and distributions using statistical tests, both over time and against a reference. The tool can automatically flag and alert on various data changes such as trends, shifts, peaks, outliers, anomalies, and changing correlations, utilizing monitoring business rules. It supports numerical, ordinal, and categorical features, and can track higher-dimensional histograms, including correlations between features. popmon provides self-contained HTML reports for easy sharing and offers integrations with tools like Grafana and Kibana for advanced dashboarding and alert handling.

HIVE Digital Technologies Ltd specializes in building and operating next-generation Tier-I and Tier-III data centers. These data centers are powered by clean energy sources, primarily hydroelectric, across Canada, Sweden, and Paraguay. The company offers a dual-engine infrastructure, combining Tier-I computing services with GPU-based accelerated AI computing. This setup delivers scalable and environmentally responsible solutions for the digital economy, supporting both Bitcoin mining and advanced AI applications. HIVE is expanding its global footprint, including a significant strategic expansion in Paraguay, and operates a fleet of thousands of next-generation GPUs to support the fast-growing AI and HPC markets.

Localeasy.io is a localization management platform designed to streamline the process of localizing software and applications. It provides free services, making localization more accessible for a wider audience. Key features include integrations with popular version control systems like GitHub and GitLab, which likely facilitate seamless workflow for developers. The platform also incorporates AI translations, aiming to enhance efficiency and potentially reduce manual effort in the translation process. Localeasy.io targets both developers and translators, offering tools to make their localization tasks more manageable and effective.