-
Our agents argued endlessly. Here's how a hybrid AI pattern tamed LLM chaos
A deep dive into building a medical ranking PoC where pure LLM reasoning failed, and how a hybrid pattern combining LLM feature extraction with a deterministic rule engine achieved stable, auditable results.
-
Vision-language model pipeline debugging: lessons from visual monitoring
Hard-won insights from building proof of concept vision-language model pipelines for visual monitoring, where hallucinations hide in plain sight, preprocessing decisions make or break everything, and the question is: can we forget classical computer vision?
-
Machine learning metrics for undefined projects: 3 critical mistakes
When building ML solutions without established playbooks, the wrong approach to metrics and validation can derail projects before you prove they work. A pragmatic framework for research, baselines, and deployment constraints.
-
Pragmatic LLM debugging: a survival guide to chaos
My approach to breaking down complex RAG and agent systems when time is short, golden datasets are missing, and quality needs a fast boost.