Contact us
Join Community
Contact us

Exciting Announcement! In celebration of launching our AI Certification, we’re thrilled to offer a 50% discount exclusively. Seize this unique chance—don’t let it slip by!

Learn more

Network Copilot – Exploring the Capabilities and Utilities of a Gen AI Assistant

AI utilized in network management involves the deployment of artificial intelligence methodologies like machine learning, deep learning, and natural language processing. Its objective is to enhance the effectiveness, dependability, and security of any network.
Aviz Network Copilot 1.0 serves as an AI-driven network analysis tool designed to assist network operators in identifying performance bottlenecks and resource utilization challenges within their networks. By leveraging natural language prompts, operators can easily interact with the tool to gain insights into network performance metrics and effectively address any issues that may arise. This intuitive approach enhances the efficiency of network monitoring and troubleshooting, enabling operators to maintain optimal network performance and reliability.

Capabilities:

Data ingestion and Storage:

It entails the procedures associated with gathering and housing data from SNMP and ONES Collector, which includes metric data collection from EOS, SONiC, Cumulus and NXOS. This encompasses acquiring, transforming, and preserving data in an organized format to facilitate subsequent analysis or utilization. Standardized data collection makes Network Copilot ready for multi-vendor on-prem deployments.
Figure-1 : Device inventory

Chat Prompt:

Users will receive a sample question to start a conversation in a chat-based interaction. This question prompts participants to respond, steering the conversation. Additionally, the chat history can be exported using this prompt.

The end user can encounter a continuous stream of data or information as a response to the question via streaming.

Figure-2 : Network Copilot homepage
Figure-3 : Network Copilot Chat window

Import Context:

In the model, network compliance is defined based on user preferences. Threshold values are adjustable to meet individual customer needs. The context can be imported and controlled through the RAG in chromaDB.
Figure-4 : Network Compliance modification
The model offers a content template for upload. Users can download it, make edits as needed, save the changes, and then upload the modified file.

Multi-Language support:

Aviz Network Copilot 1.0 will facilitate user interaction with the model in both English and Japanese languages initially. However, the capability to support additional languages can be improved based on specific requirements as needed.
Figure-5 : Conversation in English
Figure-6 : Conversation in Japanese

Analytics

Graphical representation of data using Pie Charts, Bar Charts, and Timeline Graphs to convey insights, trends, and patterns more effectively. The model also facilitates summarizing, describing, and analyzing the gathered data, while also performing computations to determine averages, counts, percentages, and more. This aids in achieving greater clarity and comprehension of your network.

Figure-7 : Pie Chart representation of inventory
Figure-8 : Line Chart representing network utilization over time
Figure-9 : Example shows average cpu utilization for past 3 months

Security:

  • Network Copilot 1.0 ensures the security of the customer data through secure certificates(HTTPS) and the Secure API.
  • The secure APIs facilitate the interaction between the Large language model and Database, allowing data exchange, task execution, and information retrieval.
  • LLMs are empowered to operate autonomously, utilizing custom-built tools to perform designated functions efficiently and effectively.

Network Copilot - Use case:

Inventory and Accounting:

Capturing comprehensive network device information, including hostname, HWSKU, operating system version, interface details, its overall capacity and also the device uptime. This helps administrators to maintain inventory records and account for network assets effectively.
Figure-10: Snapshot of model responding the device details from inventory

Capacity Planning:

This simplifies the administrator’s task by forecasting network capacity details through a comparison of available bandwidth with utilized bandwidth. It also assists network operators in designing the infrastructure required to support current and future network demands effectively.
Figure-11 : Model projecting the overall capacity and utilization of past

Anomaly Detection:

Network Copilot 1.0 is additionally trained to identify network failures resulting from sudden increases in traffic. Such spikes in CPU and memory utilization can lead to failures in various components, including control planes or links within the network infrastructure. By recognizing these patterns, Network Copilot 1.0 can help mitigate potential disruptions and proactively address issues before they escalate, thereby enhancing network stability and performance.
Figure-12 : Network Copilot detecting the peak usage of traffic
Figure-13 : Model responded with HWSKU details which has peak network usage on last 3 months

Network Compliance:

The model comes pre-configured with default compliance thresholds that establish limits for all relevant metrics captured. Network Copilot is equipped to assess whether CPU, memory, bandwidth utilization, and network packet drops comply with these thresholds by comparing observed values with the predefined limits. These thresholds are customizable by users, allowing them to be adjusted as needed.
Figure-14 : Compliance supported on Network Copilot

Conclusion:

Aviz Network Copilot 1.0 leverages cutting-edge AI capabilities facilitated by large language models. Rather than relying on conventional methods for network access, users can engage with Network Copilot to assess various aspects of their network. This includes planning network capacity, identifying instances of network failure, and verifying compliance with predefined configurations and standards. Network Copilot offers an intuitive and efficient alternative to traditional network management approaches, empowering users to gain insights and make informed decisions regarding their network infrastructure.

Unlock the Network Copilot 1.0 experience—schedule a demo on your preferred date, and let us show you how it’s done!

Share the Post:

Related Posts

Transforming AI Fabric with ONES: Enhanced Observability for GPU Performance

Explore the latest in AI network management with our ONES 3.0 series Future of Intelligent Networking for AI Fabric Optimization If you’re operating a high-performance data center or managing AI/ML workloads, ONES 3.0 offers advanced

  • Neekshitha Dyasani
Read more

Global Reach, Local Insight: ONES 3.0 Delivers Seamless Data Center Management

Explore the latest in AI network management with our ONES 3.0 series ONES 3.0 introduces a range of exciting new features, with a focus on scaling data center deployments and support. In this blog post,

  • Anbarasan Ramalingam
Read more

AI Fabric Orchestration: Supercharging AI Networks with SONiC NOS

Explore the latest in AI network management with our ONES 3.0 series As the demand for high-performance parallel processing surges in the AI era, GPU clusters have become the heart of data-intensive workloads. But it’s

  • Pramod Taramatta & Dorababu Arigi
Read more

Network Copilot – Exploring the Capabilities and Utilities of a Gen AI Assistant

AI utilized in network management involves the deployment of artificial intelligence methodologies like machine learning, deep learning, and natural language processing. Its objective is to enhance the effectiveness, dependability, and security of any network.Aviz Network Copilot 1.0 serves as an AI-driven network analysis tool designed to assist network operators in identifying performance bottlenecks and resource […]