AI reference architectures
On This Lesson 4 topics
This lesson is planned. The outline below lists the topics it will cover. The full lesson has not been published yet.
In this lesson
pick between assistant, retrieval over enterprise content, and agent with tools for a given capability.
Planned topics
- Assistant: a model answers from its general knowledge and the conversation; simple, but knows nothing about your orders.
- Retrieval over enterprise content (RAG): the system fetches relevant documents first and gives them to the model to answer from.
- Agent with tools: the model chooses actions, such as "look up order", through tools your platform exposes.
- Start with the simplest that meets the need; each step adds risk, cost, and testing.
Draft outline details
Demonstration
a three-row table of reference architectures with columns "what the model sees", "can it change data", "main risk", and a Larkspur example per row (generic FAQ bot, returns-policy answers, order-status lookup).
Check your understanding
Which architecture suits "answer from our returns policy"?
Show answer
Retrieval over enterprise content.
Practice
Apply this lesson with the module's lab, exercises and quiz.
Marking a lesson complete is optional and you can undo it any time. Your progress stays in this browser.