DiscoLike is the search engine for the business web. Google indexes webpages and returns webpages; DiscoLike indexes 80M+ business websites and returns companies, plus the right people at those companies. Go-to-market teams and AI agents describe an ideal customer in plain language and get every matching company from DiscoLike's own index, covering 98%+ of the market and roughly 3x LinkedIn's company coverage. Three ways to search: domain lookalike, natural language ICP, and exact phrase matching on homepage content. Contact search then surfaces the right people at those companies with email, phone, LinkedIn URL, seniority, department, and skills. This is the opposite of LinkedIn-derived tools. Apollo, ZoomInfo, Cognism, and Lusha start from contact records and back into companies, so buyers are searching self-reported taxonomy. DiscoLike starts from the open web, using custom LLMs trained on actual website content to match companies by what they do. Also included: technology and vendor detection across 80M+ domains, corporate hierarchy mapping, company name-to-domain matching, digital footprint scoring and growth metrics, ML-powered ICP segmentation, and exclusion lists of up to 250K company domains and 500K contacts on every plan. Four ways in, one engine: an MCP server for AI coding assistants such as Claude Code, Cursor, VS Code, and Windsurf, used by over half of active users; a REST API for Clay, N8N, Zapier, CRM enrichment, and product integrations; a web app for point-and-click TAM building; and nine bulk flat-file datasets delivered daily, weekly, or monthly, including direct to S3. Plans run from $99/mo to $1,599/mo and all include the full feature set, differing only in usage rates, included record volume, API throughput, and support. Every subscription dollar goes toward credits: each search costs a query fee plus a fee per 1,000 new records, and a record already retrieved in the last 90 days is free.