AI NetOps: How AI and Machine Learning Transform Network Operations скачать в хорошем качестве

AI NetOps: How AI and Machine Learning Transform Network Operations

AI is changing network operations (NetOps) from static automation into adaptive, data-driven systems that can summarize incidents, retrieve knowledge, and guide remediation with human oversight. In this talk, Phil Gervasi breaks down what “AI for NetOps” really means in practice, including the difference between classical ML and large language models (LLMs), why data pipelines matter more than model tuning, and how patterns like RAG (retrieval augmented generation), text-to-SQL, and agentic workflows turn raw telemetry into decisions. Topics covered: • Why AI is different from traditional network automation • Practical NetOps use cases: incident triage + summarization, knowledge retrieval, automation assistance, and AI-guided traffic analysis • The data foundation: ingest, process, store, serve (and why data engineering is the hard part) • How RAG works: embeddings + vector databases + grounded answers • Text-to-SQL: natural language analytics without memorizing query syntax • Agentic workflows: coordinating tools (traceroute, telemetry, KB) with guardrails • Challenges: hallucinations, evaluation, privacy, real-time data, and compute cost • Build vs buy: when to build in-house vs adopt a trusted platform, and why hybrid often wins Explore Kentik's ai solutions for NetOps: Kentik AI (solutions): https://www.kentik.com/solutions/kent... AI Advisor: https://www.kentik.com/product/ai-adv... Cause Analysis: https://www.kentik.com/resources/caus... Kentik Network Intelligence Platform: https://www.kentik.com/product/kentik... Kentik AI docs (KB): https://kb.kentik.com/docs/kentik-ai #NetOps #AIOps #NetworkObservability #NetworkIntelligence #LLM #rag CHAPTERS: 0:00 The Evolution of Network Operations with AI 0:35 Understanding the Complexity of Modern Networks 1:14 Advancements in Machine Learning for Networking 1:59 AI Applications in Incident Management 2:54 Building a Robust Data Pipeline for AI 3:46 Leveraging Data for Enhanced AI Responses 5:01 The Future of Autonomous AI Workflows 6:02 Navigating the Build vs. Buy Dilemma 6:39 Empowering Engineers with AI Insights 6:57 The Future of Adaptive Networking 7:23 Creating Intelligent Networks for Tomorrow