A throughput research postmortem from Espresso: what we tested, what failed, what stayed slower than baseline, and where the real architectural blockers are.
Mar 28, 2026 engineering 9 min read
Tag
2 posts tagged with llm.
A throughput research postmortem from Espresso: what we tested, what failed, what stayed slower than baseline, and where the real architectural blockers are.
If you've shipped AI features in an iOS app recently, you know the drill. Each provider has its own SDK. Switching means rewriting everything. Conduit fixes this.