IDEONYX Logo
IDEONYX

NCP (No Code Platform)

NCP (No-Code Platform) is a revolutionary platform designed to democratize the deployment and management of AI-enabled edge systems. Powered by the NVIDIA Jetson Orin NX 16G, NCP allows users to effortlessly create, train, deploy, and monitor AI models without needing extensive programming knowledge. Our proposed platform provides our customers with tools and frameworks to easily build and deploy computer vision applications. It simplifies the development process by offering a comprehensive suite of services, including pre-trained models, edge device deployment, and real-time analytics. This platform targets a wide range of industries such as manufacturing, healthcare, public transport, and retail. For example, the proposed no code platform is designed to bring advanced AI capabilities to a wide range of industrial applications. A variety of AI models based on NVIDIA capabilities are provided to perform real-time visual inspections on production lines, identifying defects or inconsistencies in manufactured products. This can improve quality control and reduce waste. We leverage the advanced capabilities of NVIDIA AI chips to boost the performance and efficiency of computer vision applications. Here’s how we achieve this.

Optimized AI Inference

The NVIDIA Jetson Orin NX is designed for high-performance AI inference with low power consumption. We utilize these features to run complex AI models for tasks such as object detection and image recognition directly on edge devices.

Hardware Acceleration

By harnessing the GPU acceleration provided by NVIDIA Jetson devices, our platform performs real-time image processing and analysis. This capability is essential for applications demanding high-speed data processing and low latency.

Scalability

Our platform software, combined with NVIDIA’s hardware, allows for the scalable deployment of computer vision applications across various environments.

Edge AI Deployment

Our support for edge deployment enables applications to run directly on our devices, facilitating real-time decision-making without the need for constant cloud connectivity.

Pre-Trained Models

We offer a library of pre-trained models for diverse computer vision tasks, including object detection, image classification, and facial recognition. This allows users to quickly integrate these models into their applications.

Real-Time Analytics

The platform supports real-time processing and analysis of visual data, which is vital for applications requiring immediate feedback, such as automated quality control in manufacturing.

User-Friendly Interface

Focusing on ease of use, our platform provides an intuitive interface to help customers quickly start building and deploying computer vision applications. The user interface provides:

In addition, there are several advantages to utilizing NVIDIA features on our platform. Our platform simplifies and presents NVIDIA features, including TensorRT, cuDNN, and CUDA, to our customers, making these advanced technologies accessible even to those who are less familiar with coding. This approach enables users to easily configure their models and take full advantage of NVIDIA’s powerful AI capabilities. To enhance user experience, our platform offers a visual configuration tool where users can drag and drop components to build and modify their neural networks. This intuitive interface allows even non-technical users to set up and configure AI models with ease. Additionally, pre-configured templates for common tasks such as object detection, image classification, and semantic segmentation streamline the setup process. These templates come with best-practice optimization techniques like precision calibration and layer fusion, automatically applied to ensure models are efficient and performant. We have developed step-by-step wizards that guide users through the optimization process, explaining each step and its impact on performance. These wizards demystify the complexities of model optimization, making it more accessible to users of all skill levels. Furthermore, our platform includes tools that automatically benchmark the performance of models before and after optimization, providing clear visual reports that highlight improvements in key metrics such as inference time and accuracy. This benchmarking capability is complemented by a comprehensive analytics dashboard that displays real-time performance metrics, helping users monitor and understand their models’ performance. To further simplify the user experience, we have designed a high-level API that abstracts the complexities of TensorRT. This API provides simple functions for loading models, running inference, and retrieving results, making it easier for developers to integrate TensorRT into their applications. The documentation includes numerous code snippets and examples that demonstrate how to use the API for common tasks, facilitating quicker learning and implementation. Our platform also ensures seamless integration with popular machine learning frameworks like TensorFlow and PyTorch by developing wrappers that enable TensorRT optimization within these environments. This compatibility allows users to leverage TensorRT’s performance benefits while continuing to use their preferred frameworks. Comprehensive documentation and tutorials further enhance the platform’s usability. We provide clear and detailed documentation that includes installation instructions, usage examples, and descriptions of each function. Tutorials and step-by-step guides walk users through common use cases, helping them get started quickly and understand how to leverage the platform effectively. Additionally, community forums and support resources are available, enabling users to share knowledge and seek assistance as needed. By incorporating these features, our platform enhances the usability and accessibility of NVIDIA’s advanced AI technologies, empowering users to build and deploy high-performance AI applications with ease.