Netscale’s AI is best understood as a recommendation + graph intelligence system, not a single “magic AI”. It combines three main AI layers: 1️⃣ Graph Intelligence (Core of the product). At its heart, Netscale works on a network graph. How it works • Every person = a node • Every connection, interaction, or shared context = an edge • Events, companies, interests can also be nodes The AI analyzes: • Distance between people (1st, 2nd, 3rd degree) • Shared attributes (industry, role, location, interests) • Interaction strength (how often, how recently) What this enables • “You should meet X at this event” • “This person is highly relevant to your goal” • “These 3 contacts can introduce you” 2️⃣ Machine Learning Scoring & Ranking On top of the graph, ML models score and rank relevance. Inputs the model considers • Profile similarity (role, industry, seniority) • Intent signals (what you want: hiring, sales, partnerships) • Event context (who’s attending, why, timing) • Past behavior (who you accepted / ignored) Output A relevance score, for example: “This person has an 87% relevance score for you at this event” This is done with: • Netscale ML model • Ranking algorithm • Collaborative filtering 3️⃣ LLMs (Generative AI layer - UX enhancement). Large Language Models to enhance the experience. What LLMs are used for • Understanding free-text goals • Summarizing why someone is relevant • Generating explanations, intros, or suggestions • Improving search and filtering with natural language Example Instead of rigid filters, the user types: “Who should I meet to sell B2B SaaS at this conference?” “I want to meet fintech founders raising Series A” The LLM: • Interprets intent • Translates it into structured filters + graph queries • Returns ranked results