Artificial Intelligence (AI) has been around for a while, having first been turned into a field of study in 1959 at Dartmouth College (a Juniper Mist customer!). Yet in the past few months, the amount of interest in AI has completely burgeoned with the introduction of ChatGPT and Large Language Models (LLMs).
While LLMs represent a huge leap forward in the field of AI, it’s dangerous to assume they replace the need for existing AI-driven virtual network assistants like Marvis. Yes, a generative conversational interface makes them extremely easy to use and extraordinarily insightful in their replies. But their answers are based on data captured during a specific snapshot in time, and they can be easily influenced to provide incorrect answers. This currently makes them ill-suited for real-time insight and troubleshooting.
The best scenario is one in which LLMs are used to augment the conversational capabilities of an existing Virtual Network Assistant (VNA). This is exactly what Juniper announced today at Mobility Field Day 9.
Optimizing User Experiences with Marvis and ChatGPT Integration
Marvis now leverages an LLM API to respond to user queries for technical documentation and other publicly available historical knowledge base information. Marvis LLM interface will also cite resources for more information. This helps IT administrators get the necessary information within an almost immediate response rate – ensuring the best user experiences possible. Here are some real-world examples:
Integrating Third-Party-Labeled Data from Zoom – Marvis Learns from More User Experience!
Juniper also announced another groundbreaking capability within Marvis designed to address a widespread problem in many enterprise networks – bad video conferencing experiences. All too often, it’s difficult to determine the root cause of dropped or pixelated calls. Is it Wi-Fi interference? WAN latency? High CPU utilization on a laptop? Or the application itself?
Marvis now integrates with Zoom to help address this situation. More specifically, Marvis pulls data from both the Zoom client and the Zoom cloud and correlates it with network-specific insights to determine the root cause of a bad video experience. In addition, Marvis learns what causes poor Zoom experiences over time, allowing it to understand trends and identify anomalies for proactive and predictive root cause identification (and correction). This gives IT teams an edge in reducing the mean time to repair Mean Time to Repair (MTTR) Zoom issues and, in many instances, can reduce Zoom support tickets entirely.
Zoom insights can be gathered using Marvis’ conversational interface. For example, as shown below, users can ask Marvis to list all Zoom users and to troubleshoot a specific Zoom session. With this integration, ensuring video calls are running smoothly and effectively has never been easier.
Juniper constantly expands Marvis to leverage new data science tools and data sources for the best user and operator experiences. LLM and Zoom are the latest in a long history of VNA advances, dating back to when Marvis was first launched in 2018. Regarding leveraging AIOps for automation, insight, and client-to-cloud assurance, Marvis continues to set the standard in the networking industry.
- Press Release: AIOps Leadership
- Blog: Access Assurance
- Marvis Product Page
- Marvis Datasheet
- AP24 Product Page
- AP24 Datasheet
- Expanded Marvis Conversation Interface with ChatGPT
P.S. We’re not magicians. We’re also not AI washers. Take a look behind the scenes to see how AI gets real.
Statement of Product Direction
Juniper Networks may disclose information related to development and plans for future products, features or enhancements, known as a Plan of Record (“POR”). These details provided are based on Juniper’s current development efforts and plans. These development efforts and plans are subject to change at Juniper’s sole discretion, without notice. Except as may be set forth in definitive agreements, Juniper Networks provides no assurances and assumes no responsibility to introduce products, features or enhancements described in this presentation. Purchasing decisions by third-parties should not be based on this POR and no purchases are contingent upon Juniper Networks delivering any feature or functionality depicted in this presentation.