Knowledge Base

A structured reference for engineers building production AI systems. Every article is written to be read once and referenced often — no fluff, no affiliate links, no sponsored content.

The knowledge base is organised into six sections covering the full stack from retrieval infrastructure to agent design to evaluation and production operations.

Sections

  • Foundations — vector databases, distance metrics, embedding concepts, and index types explained from first principles.
  • Retrieval and RAG — hybrid search, metadata filtering, embedding versioning, and end-to-end retrieval pipeline design.
  • Agents and Workflows — tool use, orchestration patterns, memory systems, and multi-agent coordination.
  • Evaluation and Quality — measuring retrieval quality, output evaluation, regression testing, and production monitoring.
  • Knowledge Graphs — graph-based retrieval, GraphRAG, and integrating ontologies with vector search.
  • Production Basics — deployment patterns, scaling trade-offs, observability, and operational cost management.