How we write the blog

Who writes it

Every post carries the name of the person at LexData who stands behind it: our engineers, our CTO, and the people who run product and go to market. Each name links to a page listing everything that person has published.

How a post is made

We choose the topics and write the brief for each post: the industry, the camera, the problem a team on site is trying to solve. Drafts are written with AI assistance from those briefs. No draft publishes as it comes out.

Before a post goes live it is checked against what our product actually does, and any claim we cannot back is cut. A number appears only when one of our case write-ups documents it, and only in the industry it came from. Headings say what the section says. A post that fails a check does not publish until it passes.

Where the pictures come from

Every post carries one frame from our own footage or a generated scene, with the boxes drawn by our models. The caption under it says which. We do not use stock photos, and no picture on the blog shows a generated face.

Why we write it

Our customers run vision models on cameras in plants, yards, substations, stores and fields. Most of what goes wrong with those models happens after launch, when the site changes and the model does not. The blog is where we write down what we know about keeping them accurate.

If something is wrong

Write to reach@lexdatalabs.com with the link and what is wrong. We check it, fix the post, and say on the post that it was corrected.