SymphonyAI has introduced an industrial large language model (LLM) designed to accelerate industrial transformation on a large scale.
The SymphonyAI Industrial LLM has been trained on an industrial dataset consisting of 3 trillion data points, over 500,000 machine tests, 150,000 components and 80,000 different assets, the provider of predictive and generative enterprise artificial intelligence (AI) Software-as-a-Service (SaaS) said in a Wednesday (Nov. 8) press release.
The Industrial LLM is hosted on Microsoft Azure and efficiently connects and contextualizes manufacturing operation information at all levels, from individual assets to global multi-plant operations, according to the release.
It serves as a self-contained intelligence source for addressing asset performance and reliability queries, the release said. By deriving insights from events, sensor data, asset details, work orders and other data sources, the Industrial LLM can provide meaningful context-aware data to operators and plant managers.
The Industrial LLM is a game-changer in an industry traditionally reliant on siloed point solutions, Prateek Kathpal, president and CEO of SymphonyAI Industrial, said in the release. Kathpal said it will be the foundation for a new generation of industrial applications and computational processes that fundamentally transform how tasks are executed and insights are derived from available data.
The Industrial LLM’s capabilities extend to machine condition diagnostics, prescriptive recommendations, answering questions on fault conditions, test procedures, maintenance procedures, manufacturing processes and industrial standards, according to the press release. It adapts in real-time to incoming data and actions, allowing it to keep pace with rapidly changing operational variables.
While the Industrial LLM is currently available for private preview, developers can sign up to build their own custom industrial applications through the Industrial LLM API, the release said. It will also be accessible in the Microsoft Teams AI Library and as a model in the Model Catalog in the Azure Machine Learning Studio. Additionally, universities and colleges can utilize it as a learning tool to train the next generation of intelligent manufacturing talent.
Some observers believe that the safest, most productive way forward for AI consists of vertically oriented AI models, or LLMs and GPT systems trained for industry-specific use cases on validated an audited data sets that are regularly retrained and updated per sector-specific guidelines, and are able to be fine-tuned by organizations using their own data, PYMNTS reported Monday (Nov. 6).
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.