{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/2175552946664e3b8935d5e96efd7b63\" frameborder=\"0\" width=\"1658\" height=\"1243\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1243,"width":1658,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1243,"thumbnail_width":1658,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/2175552946664e3b8935d5e96efd7b63-ca1d2fbcf63bd9f8.gif","duration":303.104,"title":"Building AI Agents, Knowledge Bases, Guardrails","description":"This Loom explains the AI agent development approach the author uses to deliver business impact rather than generic outputs. They report building AI agents for six months and helping around 20 clients, with most still using and maintaining the systems successfully. The process prioritizes nailing a specific workflow, building a knowledge base using generate, evaluate, distribute, and observe, including evaluations for contradictory or outdated information and guardrails for access and context. They also emphasize logging and having sweeper agents review activity to prevent unauthorized information access, and briefly mention a recently built dev agent team that is running well."}