Setting Up and Managing a Class

Setting up a Course

When the instructor first logs in to Sagepost, they are brought to their list of courses. To create a new course, the instructor simply clicks “Add Course,” which initiates the onboarding process.

Step 1: Course Details The instructor adds a course logo and the course name. A recommended naming convention includes the school and timing for easy identification, for example: "SMU Data Driven Marketing Spring 25" or "UCLA Stats 101 Fall 24."

Step 2: Initial Invitations (Optional) This step allows the instructor to optionally invite a small handful of students or administrators. This is not required, as a later step covers how to invite many students at once.

After the course is created, the instructor can navigate between all their courses by clicking on the course name in the top-left corner.

---

Adding Course Members

Course members are assigned one of two roles: Admin or Student.

Admins Cannot: Link courses, adjust billing, or adjust core course settings.

Invitation Methods

Both roles can be added either under Course Settings → Members or by clicking “Invite Member” in the bottom of the left pane.

Members can be invited in one of two ways:

1. Pasting Email Addresses
2. Sharing an Invitation Link

Invitation links can be refreshed as often as desired or deactivated if the instructor needs to restrict self-enrollment.

Managing Members

Under Course Settings → Members, the instructor can remove members from the course or change their assigned role.

---

File Management to Build the Knowledge Base

To ensure Sagepost is most effective, the instructor must build a strong knowledge base for the Retrieval-Augmented Generation (RAG) LLM to draw from. Instructors manage this information under “Files”.

Here, they can upload or remove files, arrange them in folders, and set two crucial settings:

1. Visibility: Can students view and download this file?
2. AI Indexing: Is this file included in the RAG LLM's knowledge base?

Both settings can be changed via an on/off toggle.

For example, an instructor might share a file with students (Visibility ON) but not want it indexed for retrieval (AI Indexing OFF), such as slides for an upcoming lecture where in-person attendance is preferred before students begin asking questions.

---

Handling Q&A

Please see the Q&A section of the student guide for a general overview of how Q&A works on Sagepost.

Verifying and Responding

As soon as a question is posted, the AI generates an answer. The instructor can quickly complete the question by:

• Verifying the Answer: Simply clicking the “Unverified Answer” badge.
• Adding a Custom Answer: Clicking “Add Answer” to provide their own clarification.

Either action will close the question as completed.

Regenerating AI Answers

If the instructor is not satisfied with an unverified AI answer, they can regenerate it:

1. Change Style (Optional): Select the chevron next to the unverified answer badge (e.g., choose “Simplify” if the answer is too verbose).
2. Regenerate: Click the wand icon to create a new answer based on the revision style.

Once satisfied, verifying the answer sends an update to the students. If the AI answer remains insufficient, the instructor should add their own custom answer and leave the AI answer unverified to indicate to students that the automated response is not satisfactory.

Note: If a question is already verified, the instructor must unverify it before regeneration options become available.

Editing and Deleting Q&A

Instructors can delete posted questions, or edit and delete their own responses, by selecting the vertical ellipses in the upper right-hand corner of the response section.

---

Tracking Student Engagement

Student Overview

The “Students” pane provides a quick overview of class members, along with the number of questions and answers each student has contributed. Instructors can select students and download this information in Excel or CSV format.

Tracking Metrics

In the “Engagement” section, the instructor can view the following metrics across a date range of their choosing:

• Questions posted
• Prompt tokens consumed (reflects input complexity)
• Completion tokens consumed (reflects AI answer output complexity)

By default, performance across all students is displayed, but the instructor can select an individual student for a more granular view.

Question Overview

The “Question Overview” menu item offers a quick, detailed look at the states of all questions, sorted in descending date. This helps the instructor quickly identify open questions or active follow-up discussions that still require attention.

At the top of this view, Sagepost summarizes the main themes from the questions, providing the instructor with insight into which topics students are struggling with or are most interested in.

---

Advanced Course Settings

Linking Courses

Sagepost allows the easy linking of knowledge bases from other courses, which is useful for instructors who teach the same material across multiple sections or academic years.

Courses are grouped by dragging their names into the same group. When courses are linked, any questions in one course will use the files and Q&A from the linked courses when crafting AI answers.

Important Restrictions:

• Only the owner of a course can link other courses they own.
• Admins have read-only access to the linking feature.
• Students will not see the course links.

AI Settings

Under Course Settings, the instructor can add additional configurations to the AI models:

• Customize Responses: Allows the instructor to pass short, additional instructions to the retrieval process (e.g., "never provide direct answers on the first response"). Note that these instructions are not guaranteed to override the default prompt engineering.
• LLM Providers: Allows the instructor to choose which model to use under different scenarios. While base models often perform well, the instructor may experiment with more advanced models. Note that more advanced models may consume tokens at a faster, more costly rate.

Billing and Usage Limits

Usage limits depend on the instructor’s subscription tier (noted in the “Billing“ section) or enterprise agreement.

The “Usage” section indicates the instructor's proximity to their tier capacities. AI token usage resets at the end of the subscription period,.

Email Settings

Under Settings → Account → Notifications, the instructor can change their preferences on email frequency. The default is real-time, but the instructor can also reduce the frequency to a weekly summary or turn off notifications entirely.