https://anukumar.site/writing Posts on AI agents, RAG, multi-agent systems and LLM evaluation. en https://anukumar.site/writing/llm-evaluation-llm-as-a-judge https://anukumar.site/writing/llm-evaluation-llm-as-a-judge Fri, 16 Oct 2026 03:30:00 GMT How to build an eval set from real failures, what to check with code first, the known biases of LLM-as-a-judge and how to calibrate it, evaluating agent trajectories, and running evals like tests. https://anukumar.site/writing/multi-agent-systems https://anukumar.site/writing/multi-agent-systems Fri, 16 Oct 2026 03:30:00 GMT What multiple agents actually buy you, what they cost in tokens and reliability, the patterns that work, how to write handoffs, and why to start with one agent. https://anukumar.site/writing/rag-retrieval-hybrid-search https://anukumar.site/writing/rag-retrieval-hybrid-search Fri, 16 Oct 2026 03:30:00 GMT Most wrong RAG answers are retrieval failures. How to measure retrieval on its own, chunk by structure, combine keyword and vector search, rerank, and filter safely. https://anukumar.site/writing/agentic-ai-tool-design https://anukumar.site/writing/agentic-ai-tool-design Fri, 16 Oct 2026 03:30:00 GMT Why most agent failures come from the tools, not the system prompt: naming, granularity, what tools return, error messages that teach, safe retries and hard limits. https://anukumar.site/writing/human-in-the-loop-ai-agents https://anukumar.site/writing/human-in-the-loop-ai-agents Fri, 16 Oct 2026 03:30:00 GMT Why an approval step bolted on after an agent runs doesn't protect anything, where the gate should sit, what a reviewer needs to see, and how to resume safely after a yes. https://anukumar.site/writing/multi-tenant-architecture https://anukumar.site/writing/multi-tenant-architecture Fri, 16 Oct 2026 03:30:00 GMT The three ways to separate SaaS customers' data, how each one fails, what database per tenant really costs, and how to choose.