Insights on AI, testing, developer productivity, and building better software.
LLM-generated tests often create a 'safety illusion' - coverage metrics climb while actual defect detection plummets. We break down the seven deadly patterns of AI test generation and practical strategies to fix them.
Retrieval-Augmented Generation (RAG) is the industry standard for grounding AI in reality. We demystify the pipeline - from chunking to vector databases - and explain why it's often better than fine-tuning.
Writing unit tests is one of the most time-consuming parts of software development. We explore how modern AI tools generate tests, what they can and cannot do reliably, and how to build a practical workflow.
We're excited to announce KeelTest, our VS Code extension that generates comprehensive unit tests with industry-leading pass rates. We dive deep into the architecture of our agentic pipeline.
Agentic coding is the shift from AI that just suggests code to agents that actually do the work. We break down the history, the technical breakthroughs, and the tools like Cursor and Claude Code that are changing how we build.