Topic

LLMs

6 posts

Artificial IntelligenceNovember 18, 20263 min read

When Vector Search Wasn't Enough: A Table of Contents per Document, Searched in Parallel

Our RAG assistant indexed every document and still gave answers that missed the point. What fixed it: an LLM writes a table of contents for each document at indexing time, and at question time an LLM picks sections from those contents while vector search runs in parallel. Both result sets are merged and re-ranked before the answer is written.

RAGLLMspgvector
Read
Artificial IntelligenceNovember 13, 20266 min read

Why I Call Myself an Agentic Developer: The Evidence, With Links

Everyone is agentic this year. Here is the case file instead of the claim: agentic products, the MCP tools agents use, the agent teams I run, and the numbers behind each, from 40 s to 13 s answers to a 260 person-day plan with a working foundation in two days.

Result: 105 / 166commits made with an agent

AI AgentsClaude CodeMCP
Read
Artificial IntelligenceNovember 12, 20266 min read

One Lead, Three Agents: Building an On-Prem AI Email Assistant with Claude Code, Codex, Cursor and Antigravity

A 260 person-day plan, and in two days a working foundation in 110 commits. How a spec, one rule book, a team file and a lead agent kept four coding agents in line, and how we measured the local AI instead of trusting it.

Result: 260 PD → 2 daysto a working foundation

Claude CodeAI AgentsOllama
Read
BenchmarkArtificial IntelligenceOctober 18, 20262 min read

RTX 5090 vs DGX Spark for a Local AI Agent: Why Memory Bandwidth Won

Same model, same Ollama version, same request: the RTX 5090 answered in ~15 seconds and the DGX Spark in ~41. Here is why bandwidth, not memory size, decided it, and when the Spark is still the right box.

Result: 15 s vs 41 ssame model, same request

LLMsGPUsOllama
Read
Case StudyArtificial IntelligenceJuly 17, 20263 min read

Why Our RAG Chatbot Took 40 Seconds to Answer, and the Plan to Make It 70% Faster

An end-to-end audit of a multi-tenant .NET RAG chatbot found 7-8 LLM calls per request, a duplicate vector search and 1.5 seconds of artificial streaming delay. Here's where the time went and the fix plan.

Result: 40 s → 13 stypical answer time

RAGLLMsSemantic Kernel
Read
Artificial IntelligenceMarch 10, 20261 min read

Building an Advanced RAG Pipeline for Developer Knowledge Discovery

How we leveraged LangChain and vector databases to build a context-aware search engine for internal APIs.

RAGLangChainVector DB
Read