Topic
Architecture
How the pieces of an agent runtime fit together, and who owns which one.
11 posts
- Where the seconds go in an in-app agent turnWe measured a slow turn end to end. Most of the time was not in our code and not in the model, which changes what is worth optimising.Tools and MCP10 min
- Connect an AI agent to the backend you already runYou do not need a new API surface for an agent. You need to describe the one you have, and decide what it is never allowed to touch.Tools and MCP11 min
- MCP for in-app agents: connecting your tools to a modelThe Model Context Protocol replaces one connector per backend with one protocol. What it is, how a tool call actually happens, and what it costs on the first turn.Tools and MCP10 min
- How to add an AI agent to a mobile app: the complete guideWhat an in-app agent is made of, what each of its five pieces costs to own, and the honest path from a first spoken turn to something you can hand to users.Business12 min
- How fast does a voice agent have to be? A latency playbookWhere a spoken turn's milliseconds go, what the turn-taking research actually measured, and which tactics shorten which segment of the budget.Voice9 min
- In-app agents that finish the task instead of answeringAn assistant that describes where a setting lives competes with your own navigation. One that completes the request does not. What changed, and what to build.Agent basics8 min
- What an in-app agent SDK actually doesThe boundary between an agent SDK and your app, the five phases of a turn and what breaks in each, and why the write path is the part that decides the project.Agent basics8 min
- Shipping UI without a release: the case for server-driven UIWhy mobile teams move layout decisions to the server, what it costs in versioning and coupling, and what changes when an agent is the thing choosing the shape.Render spec8 min
- The architecture of an in-app agent, for people who approve itEvery box in the system, who owns it, where the trust boundaries sit, and what changed between the intent-classification era and the one where a model chooses.Tools and MCP9 min
- Text-to-speech for agents when the language is hardSynthesis quality is decided upstream of the model that makes the sound. The two stages that break on a hard language, with Arabic as the worked example.Language9 min
- Why pre-LLM assistants failed, and what changedAssistants before 2023 broke at the layer that turned words into actions. Three things replaced it: open intent, named tool calls, and a loop that checks first.Agent basics10 min