Zain Haroon — full-stack engineer in Lahore, Pakistan. Building production web platforms with Next.js, NestJS, React and Node.https://zainharoon.comen-ushttps://zainharoon.com/blog/how-to-use-a-vector-databasehttps://zainharoon.com/blog/how-to-use-a-vector-databaseEmbeddings, similarity search and RAG — how retrieval actually works, why chunking and hybrid search decide your quality, and when a vector database is the wrong tool entirely.Sat, 11 Jul 2026 00:00:00 GMTLLMsRAGAIzainharoon890@gmail.com (Zain Haroon)https://zainharoon.com/blog/how-to-retain-contexthttps://zainharoon.com/blog/how-to-retain-contextContext windows are the working memory of an LLM application, and they fill up faster than anyone expects. What actually consumes them, why long conversations degrade, and the techniques that work.Tue, 07 Jul 2026 00:00:00 GMTLLMsAIContextzainharoon890@gmail.com (Zain Haroon)https://zainharoon.com/blog/how-llms-actually-workhttps://zainharoon.com/blog/how-llms-actually-workTokens, embeddings, attention and next-token prediction — what the machine is really doing, why it hallucinates, and what 'reasoning' is and isn't.Fri, 03 Jul 2026 00:00:00 GMTLLMsAIzainharoon890@gmail.com (Zain Haroon)