Machine Learning
Assemble Each RAG Generation Prompt from a Base Prompt Plus the Rules Each Question Needs
5 hours ago
Assemble Each RAG Generation Prompt from a Base Prompt Plus the Rules Each Question Needs
part of the generation brick of Enterprise Document Intelligence, a series that builds an enterprise RAG system from four bricks:…
PANet Paper Walkthrough: When Feature Pyramids Go Bottom-Up
7 hours ago
PANet Paper Walkthrough: When Feature Pyramids Go Bottom-Up
I wrote about the FPN (Feature Pyramid Network) architecture [1], which is one of the most influential necks we can…
Setting Your Own Great Language Model
1 day ago
Setting Your Own Great Language Model
: Frontier AI models are increasingly at risk of being trapped behind strict export controls or mounting API costs. As…
Stop Returning Text from RAG: The Typed Answer Contract That Prevents Hallucination
1 day ago
Stop Returning Text from RAG: The Typed Answer Contract That Prevents Hallucination
brick of Enterprise Document Intelligence, a series that builds an enterprise RAG system from four bricks: document parsing, question parsing,…
AI Agents Explained: What Is the React Loop and How Does It Work?
2 days ago
AI Agents Explained: What Is the React Loop and How Does It Work?
In my last post,. Calling a Tool is a method that allows an AI model to decide which function to…
Long Context vs. Short Context Model: When Does a Long Context Model Win?
2 days ago
Long Context vs. Short Context Model: When Does a Long Context Model Win?
1. 1.1 The marketing claim, and the question it skips Each new generation of encoder models comes with a bigger…
LLM Wikis Are Over-Engineered — I Replaced Mine With a Pure Python Compiler
2 days ago
LLM Wikis Are Over-Engineered — I Replaced Mine With a Pure Python Compiler
TL;DR I pipeline that compiles a folder of raw, messy text notes into a linked, linted markdown wiki. No LLM…
Untaught Lessons of RAG Retrieval: Cosine Is Not a Basis
2 days ago
Untaught Lessons of RAG Retrieval: Cosine Is Not a Basis
friend to Enterprise Document Intelligenceseries whose philosophy is set in Grow the Expert. It zooms in brick 3 (retrieval) of…
Tokenminning: How to Get More from Your Chatbot for Less
3 days ago
Tokenminning: How to Get More from Your Chatbot for Less
virus spreading through big tech. Engineers are being judged, directly or indirectly, by how much AI they can consume. More…
Amaluphu Wokuklama, Hhayi Ukwaziswa | Mayelana neSayensi Yedatha
3 days ago
Amaluphu Wokuklama, Hhayi Ukwaziswa | Mayelana neSayensi Yedatha
“Asisayibhali imiyalo. Sidizayina izihibe.” – othile e-Anthropic ngoJuni 2026 i-agent loop, ukuzigxeka akwenzanga ngcono kunokungenzi lutho. Isiqinisekisi esinqunyiwe, esinehange emthonjeni…