Explore cutting-edge search and AI technologies through this collection of lightning talks from Haystack US 2025. Discover innovative approaches to binary quantization, RAG systems without vector databases, and context optimization in retrieval-augmented generation. Learn about funding strategies for search improvements, web-scale retrieval for large language models, and agentic query understanding using search MCPs. Examine quantization design patterns and key metrics, metadata-aware vector embedding models, and ethical considerations in search and AI. Gain insights into Apache Community contributions, IoT data processing, Lucene University developments, and practical questions about search engine necessity. Each presentation offers focused expertise from industry practitioners and researchers, covering both technical implementations and strategic considerations for modern search and AI systems.