Writing
6 posts, grouped by how finished they are rather than by date. Grouping by lifecycle is the point: on most publications everything is presented as equally settled, and the reader has no way to know how much weight a claim has earned.
Evergreen
Actively maintained. Someone is answerable for the accuracy.
A field guide to evaluating retrieval systems
Recall@k tells you if you found the chunk. It tells you nothing about whether the chunk was useful.
Cache invalidation is a distributed systems problem
The TTL is not a cache policy. It is a probability distribution, and somebody has to own it.
Budding
An argument forming, with references attached to its claims.
Why your embeddings are worse than you think
Dimension buys separation, not meaning. The curve flattens hard, and three pipeline bugs were costing more than the model.
The p99 is a lie you tell yourself
You cannot add percentiles. Every latency budget that does is a budget nobody is actually keeping.
Seedlings
Published before they are finished, on purpose.
We deleted our staging environment
Six months in. The best thing we did was delete a database nobody had ever tested against.