Descriptions:
Swyx, co-founder of the AI Engineer conference series, delivers a keynote in London chronicling how a nine-person team generates over $9 million in annual revenue by embedding AI agents throughout their operations. Drawing on three years of keynotes that tracked AI’s falling cost curve (roughly 100x per 12–18 months) and the rise of “tiny teams,” this talk focuses on the concrete workflow changes that turned those trends into operational reality.
The most striking examples involve non-technical staff rather than engineers. A designer who previously waited weeks for development turnarounds now pushes changes directly via Devin with near-immediate results—and has started shipping animations and polish work that never got prioritized before. A viral tweet became a live website feature in minutes. Figma mockups became pixel-perfect production code without developer bottlenecks. Conference logistics—coordinating 130 speakers, dozens of sponsors, and hundreds of attendees across spreadsheets—are now handled through AI-assisted workflows that replaced the team’s prior CMS setup. The actual tech stack (Devin, Figma, React, Supabase, Tido, Google Sheets, Sessionize) is named throughout.
Swyx’s central argument is that the real productivity unlock is not raw code output—it is eliminating the blocking dependencies that prevent employees from acting on their ideas. He frames 2026’s defining theme as “agents for everything else”: moving beyond coding use cases toward agents that accelerate creative work, event operations, and team coordination. The result, he argues, is more creative output from existing employees, not just faster software shipping.
📺 Source: AI Engineer · Published May 01, 2026
🏷️ Format: Workflow Case Study
