{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/89efffb77c00453897d560fb03eb46b1\" frameborder=\"0\" width=\"1662\" height=\"1246\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1246,"width":1662,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1246,"thumbnail_width":1662,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/89efffb77c00453897d560fb03eb46b1-7b6537fec525114d.gif","duration":2226.624,"title":"Dinner Buddies by Tobias Kromke","description":"This Loom explains how an AI agent helps the presenter and his wife quickly decide where to eat in Singapore when dinner plans get delayed. They typically go to a mall near their home opposite their house 3 to 4 nights a week, but on nights her leaving time can shift from about 6.30 to later due to colleagues stopping her, forcing last minute decisions. The agent uses deterministic orchestration and Python to handle time math and restaurant filtering, while an LLM selects and explains recommendations based on taste. The presenter built it in about four hours and uses a curated list of 11 real restaurants with mood tags and personal notes, plus demo testing with late arrivals; a key lesson was to make the UI defensive because LLM outputs can return null backup picks."}