Show HN: I made semantic search engine for engineering blogs and conferences (devblogs.sh)

A maker on Hacker News launched a semantic search engine called DevBlogs that indexes engineering blogs from top teams, conference talks, and vetted indie writers — organized by meaning rather than keyword matches. The indexed sources include engineering sites from companies like Airbnb, Netflix, Stripe, Lyft, Uber, Jane Street and many conference talks and community blogs, aiming to surface conceptually relevant posts even when authors use different wording. The pitch: find architecture patterns, performance tips, or debugging case studies across dozens of high-quality engineering publishers without having to guess the right search terms. For the AI/ML community this matters because semantic search — typically implemented with dense embeddings + vector similarity (often from transformer encoders) and a vector database — turns scattered, jargon-rich engineering knowledge into retrievable concepts. That enables faster research, trend spotting, and easier R&D reproducibility by linking design decisions across projects. Practical implications include the need for quality crawling, document chunking and transcript handling, regular re-embedding for freshness, and careful ranking/curation to avoid bias. Paired with summarization or RAG workflows, such an index can become a powerful research assistant for engineers seeking real-world patterns and implementation details across industry-grade sources.