12 posts · rss

Microsoft 365 API Performance

Making Microsoft 365 APIs faster and cheaper: batching, caching, ETags, change detection, and what a wasted API call actually costs you.

This is the topic I care about most, because it is the one that shows up on an invoice. Microsoft 365 APIs are metered, throttled, and slower than local code, so the difference between a naive implementation and a careful one is not style. It is minutes of user time and real money.

The patterns repeat across every service: batch instead of looping, ask for less, cache what does not change, and check whether a write is necessary before making it. One job here went from ninety to a hundred and twenty seconds down to twenty-five to thirty-five. A separate one turned out to be spending most of its calls writing values that were already correct.

I try to publish the measurements rather than adjectives. Read them as evidence that a technique helped in one specific case, not as a benchmark.