{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/0094d5bb49234ab0935eb84752424844\" frameborder=\"0\" width=\"1728\" height=\"1296\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1296,"width":1728,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1296,"thumbnail_width":1728,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/0094d5bb49234ab0935eb84752424844-8eb4a057b57f77b1.gif","duration":606.119,"title":"Building a Production-Ready Semantic Search Engine from Scratch 🚀","description":"Hi, I'm Ace, and in this video, I walk you through the Defector DBQ, a production-ready semantic search engine I built from scratch for Stack.ai. I designed it with a four-layer domain-driven architecture for flexibility and scalability, implementing two indexing algorithms: flat indexing for smaller datasets and IVF indexing for larger ones, achieving around 100 milliseconds search time. I also introduced a read-write lock system to handle concurrent requests efficiently. If you have any questions or feedback, please reach out—I'm happy to discuss!"}