What a chatbot actually is
A practical system definition—from the message box to the policies, tools, people, and failure paths behind it.
Read guide ↗Field 01 / Foundations
Before choosing a model, decide what the service is allowed to know, say, and change. These guides separate product scope from interface polish and turn a vague assistant idea into an accountable service.
The question to keep asking
Map the channel, identity, state, evidence, tools and human owner.
Make supported jobs, limits and alternative routes visible.
Move verification into an appropriate secure surface.
See the source / related interface context
Interoperability and authorization are different layers.
Open the case file ↗This source capture illustrates a related interface—not evidence that the vendor passed these review checks.
The complete foundations desk
A practical system definition—from the message box to the policies, tools, people, and failure paths behind it.
Read guide ↗Choose the least flexible mechanism that can safely complete the job.
Read guide ↗
Ubisoft · source visual / reviewFoundations2 minDesign session state, user profile, retrieval context, and long-term memory as separate promises.
Read guide ↗Web chat, email, voice, messaging, and kiosks change what good behavior means.
Read guide ↗Publish what the service can do, where it stops, and the next route a person can actually use—then test refusals as carefully as answers.
Read guide ↗A first run should reveal useful tasks, limits, and exits without turning a new user into a captive tutorial subject.
Read guide ↗Preserve owned records, documented tool contracts, and asset rights so leaving a provider is a tested transition rather than an emergency assumption. This field note supplies an inspectable artifact, counterexample, and release checks.
Read guide ↗Keep the system connected
Conversation design Knowledge & actions Put the reading to work