latent — AI systems, logged A working log on building AI systems that survive production — RAG, tool-using agents, and honest evals. By Chirag Singhal. https://ai-blog.oriz.in/ 2026-08-11T00:00:00.000Z Chirag Singhal https://oriz.in/about/ whyiswhen@gmail.com Astro Evaluating LLM outputs without fooling yourself https://ai-blog.oriz.in/blog/evaluating-llm-outputs/ 2026-08-11T00:00:00.000Z 2026-08-11T00:00:00.000Z Chirag Singhal Vibes are not a metric. A practical framework for measuring whether your LLM feature is good — grounded in test sets, rubrics, and the honest limits of LLM-as-judge. MCP: giving language models real tools https://ai-blog.oriz.in/blog/mcp-real-tools-for-llms/ 2026-08-08T00:00:00.000Z 2026-08-08T00:00:00.000Z Chirag Singhal The Model Context Protocol turns an LLM from a text generator into something that can read your files, call your APIs, and act. Here is what it is and how to build a server that does not misbehave. RAG pipelines that actually work in production https://ai-blog.oriz.in/blog/rag-pipelines-that-work/ 2026-08-11T00:00:00.000Z 2026-08-05T00:00:00.000Z Chirag Singhal Most RAG demos fall apart on real corpora. Here is the retrieval stack that survives messy documents, ambiguous queries, and users who paste in nonsense.