A recent survey conducted among IT and networking professionals has revealed that a significant majority (97%) view the incorporation of artificial intelligence (AI) and machine learning (ML) for automation within software-defined wide area network (SD-WAN) environments as a vital aspect, with many considering it essential. This research indicates that AI technologies are expected to enhance automation and operational efficiency in the increasingly complex landscape of SD-WAN.
The survey, which included 374 respondents from organizations in the US and Canada involved in networking technology, was carried out by TechTarget’s Enterprise Strategy Group (ESG). The findings suggest that as IT infrastructures become more distributed and intricate, the dynamics of SD-WAN environments will also need to adapt and evolve over time.
Data from the survey highlights that network operations teams acknowledge the necessity for a more proactive approach, aiming to improve key metrics such as mean time to detect (MTTD) and mean time to repair (MTTR). AI, ML, and automation are anticipated to play a critical role in this transformation. Among the networking professionals surveyed, 40% emphasized the importance of detecting anomalous activities, while 39% pointed to predictive analytics for early detection of issues and another 39% highlighted accelerated troubleshooting as essential features for their SD-WAN frameworks.
AI is expected to be utilized for providing actionable recommendations, performance optimization, and, as it gains trust, automating remediation processes without the need for manual intervention. The analysts at ESG noted the increasing risk associated with an expanded attack surface, underscoring the positive potential for organizations to utilize their SD-WAN capabilities for quicker issue detection.
Network Equipment Providers Integrate AI Capabilities
As networking professionals recognize the advantages that AI can bring to network operations, equipment vendors have been actively integrating AI and ML into their offerings, extending artificial intelligence for IT operations (AIOps) to enhance network operations.
For example, Cisco introduced its AI-driven networking, security, and observability solutions during Cisco Live 2024. This initiative aims to equip businesses with the visibility and insights necessary to secure and streamline their digital infrastructures while fostering resilience. Cisco claims that its offerings are uniquely designed to transform how infrastructure and data connect, safeguard businesses, and address core challenges faced by clients of all sizes.
In another instance, Extreme Networks launched AI Expert in April, which leverages data from various sources to boost performance and operational efficiency. The AI Expert platform integrates data from applications and devices throughout the network to establish performance insights. It automates operations and generates alerts for issues like network congestion, degradation, or Wi-Fi dead zones, thereby curating enterprise data for actionable insights.
Juniper Networks also unveiled a new AI-driven product in June, designed to enhance enterprise WAN environments and deliver assured SD-WAN experiences with proactive AIOps. Marvis Minis, Juniper’s digital experience twin, has been enhanced for SD-WAN to diagnose authentication problems without user involvement, allowing continuous WAN speed testing and proactive issue resolution.
Transforming Network Administration with GenAI
Generative AI (GenAI) represents a promising frontier for additional applications beyond AIOps in networking. The industry is currently facing a notable IT skills crisis, as John Burke, CTO at Nemertes Research, noted in a piece published on SearchNetworking. He highlighted that constrained IT budgets contribute to this skills gap, with new IT professionals often prioritizing broader IT skills over specialization in networking.
Instead of seeking to hire experienced network managers, organizations could leverage GenAI to assist less experienced IT staff or those outside the networking domain in effectively managing networks. Burke believes that as GenAI technology matures, it could automate routine tasks, enhance incident response, and help bridge the workforce gap, serving as a supportive tool for network administrators.
By employing GenAI, network teams could use its natural language capabilities to simplify documentation processes. For instance, GenAI can analyze existing configuration files and network maps to produce comprehensive written descriptions, even generating diagrams when appropriate. If integrated alongside machine learning tools, GenAI could empower teams to manage increasing workloads more efficiently, even amid declining staffing levels.
Burke envisions applications where network administrators could articulate their network intentions verbally, and GenAI could generate necessary commands for implementation, as well as offer descriptions of the network’s configurations. Additionally, GenAI could aid in auditing configurations, providing significant support to IT teams.
Beyond networking, GenAI could assist programmers by aiding in code generation, providing syntax checks, and offering constructive feedback. However, Burke cautions that network engineers must still critically evaluate and finalize any code generated by GenAI before deployment.
The Future of AI in SD-WAN
The introduction of AI-enhanced functionalities in network administration tools highlights a growing awareness within the industry regarding the intricacies of modern corporate networks that support distributed IT ecosystems. These advanced solutions hold the promise of making such networks easier to manage.
However, as networking complexity is likely to increase, professionals in the field will encounter greater demands. While experts do not foresee the emergence of entirely automated network management, the assistance provided by AI tools is expected to be invaluable.