Notes on RAG (Part 2): Hybrid Search, Reranking, and Evaluation
A continuation on moving from baseline to production with hybrid retrieval, reranking, contextual retrieval, and a practical evaluation loop.
AI Builder and Data Storyteller
Notes on AI systems, environmental modeling, and applied data science.
A continuation on moving from baseline to production with hybrid retrieval, reranking, contextual retrieval, and a practical evaluation loop.
A conceptual introduction to why RAG matters, how baseline retrieval works, and what early failure patterns to expect.
A technical framework for evaluating LLM systems with deterministic checks, model based judges, retrieval metrics, and human review loops.
A practical PINN walkthrough in PyTorch that models water tank drainage by enforcing Torricelli's law directly in the loss function.
An applied tutorial on creating a recommendation workflow with retrieval-augmented generation and vector search.
A practical look at groundwater flooding in Bavaria and how Bayesian inference can make uncertainty visible in risk maps.