Golioth for AI Streamlines IoT Data Transforms, ML Models, and Inferencing on Edge Connected Devices
Partnership between Golioth AI and Edge Impulse streamlines capturing data, building an AI model, and updating it over-the-air.
Golioth announced a significant feature launch called Golioth for AI. This new capability is a comprehensive suite of features designed to simplify and enhance the integration of AI into IoT products. The new features focus on three core areas: training data, model management, and inference.
Golioth is an IoT platform that provides cloud services for embedded devices. In short, it simplifies the messy process of connecting a microcontroller to the Internet by delivering device-level services like messaging, security, and over-the-air (OTA) updates.
Golioth for AI merges AI with IoT devices. It enables engineers to leverage the rich data IoT devices collect to train and create better AI models. In addition to sensor data like position or temperature, Golioth supports Pipelines that accommodate large payloads like audio or image data. These uploads can go to object storage destinations like AWS S3, Azure Blob Storage, and
Another substantial announcement is Golioth's partnership with Edge Impulse. This platform allows engineers to quickly build datasets, train models, and optimize libraries to run on edge computing devices' microcontrollers. This partnership facilitates sending IoT data from Golioth to Edge Impulse and leverages the strengths of both platforms to streamline AI model training, especially for microcontroller-class devices.
The over-the-air (OTA) update system now supports an expanded set of artifact types, such as AI models and media files. This improvement is noteworthy because AI models can be updated independently of firmware updates, reducing bandwidth, battery consumption, and downtime.
Golioth for AI offers robust support for on-device and cloud inference methods, allowing engineers to choose the best approach for the application's needs. On-device inferencing is ideal for real-time monitoring or offline operation, while cloud inferencing is suitable for tasks requiring significant processing power. Golioth's Pipelines further enhance this flexibility by supporting transformations and destinations, seamlessly integrating devices with AI platforms like Hugging Face, OpenAI, and Anthropic.
Whether you're an AI beginner or an expert, Golioth's platform offers the tools and resources necessary for advanced AI integration with IoT. For more information and to get started with examples, check out this blog post, Introducing Golioth for AI.
