Understand a system
See a domain as an operational whole before following its implementation seams.
LKCI Engineering Field Guide
Explore source-backed architecture stories, reusable operating patterns, and field notes from building composable systems.
Choose a question
See a domain as an operational whole before following its implementation seams.
Trace how intent moves through automation, review, failure, and repair.
Find durable boundaries for data, services, jobs, integrations, and UI surfaces.
Review source-backed context and the decisions that support an architectural claim.
engineering-practice · field-note
This week started as a website project and ended as a lesson about coordination.
precon · case-study
Files arrive before meaning does.
platform · case-study
Some failures mean the software is wrong. Others mean the software has reached the edge of what it can decide safely.
data · case-study
One Postgres database can make several kinds of truth look interchangeable. They are not.
platform · case-study
Capability is getting cheaper. Authority is not.
geo · case-study
A map can reveal a pattern and still fail to preserve a decision.
platform · architecture
Layering is how we keep a convenient change from becoming a permanent coordination problem. Runtime surfaces, stateful capabilities, deterministic policy, provider translation, data movement, executable jobs, infrastructure, and durable…
platform · concept
An operating company experiences architecture through handoffs. Does context survive them? Can the next person see what happened? When something breaks, is there evidence and a safe way forward? Those questions matter more than whether…
platform · concept
Automation is only useful to an operating company if it can run, explain what happened, and hand uncertainty to a person without losing the thread. We use four recurring shapes to make that possible: pipelines maintain data, jobs initiate…
platform · concept
A recoverable failure is unfinished business. If the company still owes a decision, a correction, or a follow-up, the system needs to preserve that fact as structured operational state. An exception can stop a process. It cannot, by…
execution · reference
A job becomes fragile when the only complete account of its behavior is the script that happened to run it. Jobs should make work executable and legible. They should not become the private home of business policy, provider behavior, and…
services · concept
If a capability exists only inside the caller that first needed it, the company will pay to rediscover it at every new surface. Services give durable ownership to behavior that must survive across APIs, jobs, UI workflows, and agents.
integrations · concept
A company cannot coordinate reliably if every internal workflow has to relearn the eccentricities of each external system. Providers will send incomplete payloads, change enums, and disagree about what the same business fact looks like.…