Network operations are the backbone of every successful enterprise. Managing these complex networks feels like navigating a maze of fragmented data, manual processes and endless troubleshooting. Network operators grapple with data islands, forcing them to manually piece together information from disparate sources through tedious processes referred to somewhat derisively as ‘click ops’. This not only consumes valuable time but also introduces a higher risk of errors and delays in identifying and resolving critical issues.
The Evolution of Network Operations: Introducing Network Copilot
Instead, imagine a world where you could simply ask your network questions in plain English and receive instant, correlated insights. This is the promise of Aviz Network Copilot, an innovative AI solution designed to revolutionize network operations by providing a natural language interface to interact with network data. It’s not about replacing existing infrastructure but enhancing it, bridging critical gaps, and empowering network professionals to work smarter, not harder.
Recently at Networking Field Day 38, Aviz showed off Network Copilot and how it can assist operations teams with solving the challenges faced with a mountain of data and no clear way to scale it.
Bridging Data Silos and Unifying Data Access
The first major hurdle in effective network management is fragmented data. Network Copilot directly addresses this by acting as a universal translator for your network’s data. It connects to existing data sources like Cisco Catalyst Center, Nexus Dashboard, Elastic, IP Fabric, Splunk, and Zend, along with direct device connections, without requiring you to rip and replace your current data lakes.
This seamless integration allows users to ask questions in native language and instantly pull together correlated data from various tools. For instance, instead of individually checking packet drops on one device and then cross-referencing ticket summaries in another system, you can simply ask the Copilot for a holistic view. This capability dramatically speeds up investigations, providing a unified context that was previously only achievable through painstaking manual effort.
This kind of querying capability means your operations teams don’t have to think in arcane syntax to get answers. They also don’t have to remember the specifics of an individual platform. Instead, by training them to ask questions in simple terms you can generate higher quality answers and ensure that terminology doesn’t create gaps in the results.
Automating the Mundane, Enabling the Strategic
Repetitive tasks are a drain on resources and a common source of operational inefficiency. Network Copilot excels at automating these essential functions. A perfect example is generating compliance reports. Network Copilot can securely retrieve configurations from devices, validate them against predefined compliance requirements (such as security features or logging standards), and then generate a customized report. For large networks or during quarterly security audits, this automation translates into significant time and cost savings. It frees up skilled engineers from tedious data collection, allowing them to focus on more strategic initiatives.
It also ensures that auditors can’t surprise you since the reports are so easy to generate. Being able to produce any kind of report on a moment’s notice means that you have complete visibility into your network and, at least according to the auditors, you aren’t making up any of the data, since they can see the report being built in real time.
Enhancing Troubleshooting and Root Cause Analysis (RCA)
Root Cause Analysis (RCA) is an area where Network Copilot truly shines. Complex network issues often require sifting through mountains of log files, console outputs, and contextual information from multiple devices. The Copilot, powered by a Large Language Model (LLM) and its agents, can parse this data, summarize key events, identify error messages, and provide potential root causes and actionable suggestions for addressing issues.
Consider a scenario where you upload log data from a problematic customer environment. The Copilot can quickly summarize device statuses, pinpoint interface issues (like status changes from up to down), detect MAC address-related problems, and offer guidance on potential issues based on its pre-trained knowledge. It goes beyond a simple event log. It aims to provide intelligent insights, saying, “this all seems to have started here because this event happened,” which is the very essence of effective RCA.
The Power of Private AI and Cost-Effectiveness
Network Copilot’s core functionality relies on a sophisticated Large Language Model (LLM) and a set of agents. The LLM translates natural language questions into data retrieval actions, while agents perform reasoning, summarization, and data manipulation. It’s designed as a private AI platform that users deploy in their own environment. This ensures that your data remains local and secure, and you maintain complete control, without training external LLMs. The system uses supervised fine tuning on a pre-trained foundation model, allowing it to understand networking-specific terminology and use cases deeply. It intelligently knows how to access and interpret dynamic operational data, handling data normalization and correlating information from different sources even if they use varying attributes, such as the difference between hostnames and IP addresses.
Bringing It All Together
Network Copilot is designed to automate the boring stuff, potentially leading to significant cost savings by reducing the need to hire specialized data scientists or invest in expensive cloud-based AI infrastructures. The solution is positioned to pay for itself within a year through improved operational efficiency and reduced manual effort.
By bridging data silos, automating repetitive tasks, and providing intelligent assistance for complex troubleshooting and root cause analysis, Network Copilot empowers network operations teams to move beyond reactive fire-fighting and towards a more proactive, efficient, and collaborative future.
For more information about Aviz Networks and their platform, make sure to check out their website at https://AvizNetworks.com. To watch more of their presentation from Networking Field Day 38, make sure to check out their presentation page here.
