Chinese AI Models: What Happens to the Data You Send Them
Servers, retention, training and the law behind DeepSeek, Qwen and Kimi. What really happens to a project document pasted into a Chinese model.
Tag Step by Step
Step-by-step guides for engineers: from the first click in a new tool to a finished workflow you can repeat in your own office.
Many Mailboxes in One System: Thunderbird and AI
A case study: several mailboxes in one Thunderbird profile, with AI triage, filters and a searchable archive instead of five open tabs.
RAG in Obsidian: Index Your Knowledge Base and Talk to Your Notes
Retrieval over your own notes: what the pipeline is made of, three ways to run it and the mistakes that make an answer sound confident without a source.
Toggl Track from Scratch: Time Tracking in a Design Office
A timer, projects, tags and reports in one working week. How to start tracking time in a design office without turning it into a second job.
Revit Performance Tuning for Large MEP Models
A large MEP model is usually slowed down by views, links, families and the workstation, not by Revit itself. The order of checks that actually helps.
