The first form extracts perfectly. Then the address moves, the label changes, or the next document calls it something else. We add another rule. This series explores how to retain what those exceptions teach us: the structure we want, the evidence that…

StyloBot Release Series: Learning to Get Faster

Heads up: this one is for the nerds. It's a deep technical dive into StyloBot's adaptive learning system (EWMA updates, hysteresis thresholds, verdict caching, variance watchdogs), with real code from Mostlylucid.BotDetection and pointers to the equivalent…

Deduplication of Graphitic RAG Evidence Segments in lucidRAG

NOTE: This is not a conventional blog article. It is a technical spec written for a concrete feature in lucidRAG. I iteratively feed this document to code-focused LLMs during development to reason about trade-offs, validate assumptions, and converge on a…

GraphRAG Part 2: Minimum Viable GraphRAG

In Part 1, we explored why GraphRAG matters. Now let's build a minimum viable GraphRAG with three extraction modes: Mode LLM Calls Best For Heuristic (default) 0 per chunk Fast indexing, structured markdown Hybrid 1 per document Balance of speed and quality…

GraphRAG: Why Vector Search Breaks Down at the Corpus Level

Your RAG system is great at "needle" questions: retrieve a few relevant chunks and synthesise an answer. It struggles with two common query types: Sensemaking: "What are the main themes across this corpus?" Connective: "How does X relate to Y across different…

Cooking with DiSE (Part 4): Building Systems That Learn, Adapt, and Evolve

Most software architectures assume the system is static. DiSE assumes the system is alive. Note: This is Part 4 in the "Cooking with DiSE" series. See Part 1, Part 2: Graduated Apprenticeships, and Part 3: Untrustworthy Gods for background. This article is the…

RAG for Implementers: CPU-Friendly Semantic Search with ONNX and Qdrant

📖 Part of the RAG Series: This is Part 4a - core implementation: Part 1: RAG Origins and Fundamentals - What embeddings are, why they matter Part 2: RAG Architecture and Internals - Chunking, tokenisation, vector databases Part 3: RAG in Practice - Building…

"Isn't DiSE Just Voyager?" - Why Structure Beats Brilliance

My system, DiSE, is a self-optimizing, self-assembling, software engineering grounded workflow builder. It generates testable code artifacts, evaluates them objectively, and evolves them over time without constant human supervision. But here's the thing:…

RAG for Implementers: Hybrid Search and Automatic Indexing

📖 Part of the RAG Series: This is Part 5 - production integration patterns: Part 1: RAG Origins and Fundamentals - What embeddings are, why they matter Part 2: RAG Architecture and Internals - Chunking, tokenisation, vector databases Part 3: RAG in Practice -…

RAG in Practice: Building Real-World Applications

In Part 1 of this RAG series, we covered the fundamentals of Retrieval-Augmented Generation—what it is, how it works, and the underlying technology (embeddings, vector databases, LLM internals). Now it's time to put that knowledge into practice. This article…