Practical AI in Cybersecurity: How CrowdStrike, Vectra, and Cloudflare Use AI

Surge of threats, automation of attacks, lack of staff, how AI eases the work of SOC, and key AI security points worth explaining to the client.

• Overview of Charlotte AI: concept and principles of use, features of Charlotte AI, prompts, prompt library and logging; • Charlotte AI and Fusion SOAR: intersections and possibilities; • Charlotte AI and query creation: IT queries and CQL queries; • Investigation of detections and incidents using Charlotte AI: analysis of detection and response to detections, principles and capabilities.

• AI prioritization of hosts (Threat, Certainty, Observed Privilege): ranking hosts by risk and determining who to respond to first; • AI host legend: attack categories/phases, connection between detections, context and key activity indicators; • AI triage of detections: statuses and AI-filtered logic, impact on display and prioritization of detections; • AI integrations with external systems: using AI and scoring in interaction with third-party systems for enrichment, correlation, and automated actions.

• Cloudflare Workers AI: deploying AI models on Cloudflare’s global edge network, supported models (LLM, image generation, embeddings) and serverless approach without infrastructure management; • Cloudflare AI Gateway: centralized control of AI traffic, request monitoring, expenses and latency, caching, rate limiting, fallback between providers; • Practical use cases: chatbots and RAG systems, content moderation and cost optimization on APIs of different LLM providers.

Answers to your questions and practical recommendations from experts.