AI engineering, the way it happens day to day.
Short articles by Lucas Palhares Barbosa, founder of Weft, on what works, what breaks and how to measure it.
Subscribe via RSS- How to measure whether an AI assistant gets it rightA test with real questions before putting the assistant in front of anyone. The score did not change, and that was the most useful measurement of the project.Read the article →
- The ROI of generative AIROI is the return on investment. Where it shows up, where it disappears, and why measurement has to be born with the project.Read the article →
- RAG fails in the plumbing, not in the modelWhen the assistant gets it wrong, the problem is usually in the path the data takes to reach it. Two real cases from one of my projects.Read the article →
- AI Engineer, ML Engineer and data scientistThree roles that often get mixed up when building a team. Working with AI is not always the same job.Read the article →
- RAG or GraphRAG: when each one is worth itTwo ways for AI to consult company documents before answering. The choice is not about technology, it is about architecture.Read the article →