Juniper Networks recently expanded the functionality of its virtual AIOps assistant Marvis to include ChatGPT. In addition, Marvis now also helps with questions about issues surrounding the Zoom video conferencing platform through an integration with Zoom-data.
The Marvis virtual assistant is supported via the Mist AI engine. This in itself already uses NLP technology. However, it now offers users even more functionality for querying and finding answers to network problems through the integration of ChatGPT.
According to Juniper Networks, interactions with the virtual assistant have now become even more “human. This is especially true when it comes to finding documentation and information about problems that require technical support.
More specifically, the Marvis assistant now uses an LLM API. This API allows the assistant to respond to user queries around technical documentation and other publicly available historical knowledge base information. This information is presented entirely in ChatGPT style, including links and relevant documents.
Zoom data now also available
In addition to the integration with ChatGPT in Marvis, the underlying Mist AI engine is now also fed with external cloud-based data on the use of the video conferencing platform Zoom. This should further extend the existing rich offering via telemetry, information and operating status from routers, switches, firewalls and access points in Mist AI.
By combining Zoom data with that from other network functionalities, virtual assistant Marvis now has a deep learning model that can accurately predict Zoom’s user experience. Using AI and ML techniques, this allows the virtual assistant to more quickly identify the root cause of video conferencing problems, answering administrator questions.