Text Chatbot

The text chatbot is the lightweight, type-only experience — no camera, no microphone, no avatar video. Choose this mode when you're studying in a quiet space (library, lab, classroom) or when you'd rather have a written transcript than a spoken conversation.

Sending a Message

Type your question in the input box at the bottom.
Press Enter to send (or click the send button). Shift + Enter inserts a newline.
Dr. Olivia's response streams in token-by-token in real time.

Session Panel

On the right of the chat is the session panel. On a phone or a narrow window, open it with the tab on the right edge of the screen. On a wider screen you can hide it with the button at its top to give the chat the full width, and bring it back with the same tab on the right edge. The text chatbot remembers your choice in this browser.

Study plan (top) — your active plan. Press Work on this plan or double-click it, and Dr. Olivia picks it up. See Study Plans.
Files (bottom) — every file she can see. Hover over a file and press the paperclip to attach it to your next message, or the × to put it away in your archive. The box icon opens the archive. See Uploading Documents.

Files attached to the message you are writing are shown just above the message box. The × there only detaches a file from that message. It stays in your files.

Markdown Rendering

Dr. Olivia's replies use full markdown — headings, bold, italic, lists, tables, code blocks, blockquotes. The text chatbot renders all of it inline.

Code blocks get syntax-highlighted styling and a monospace font.
Tables render as proper HTML tables.
Links open in a new tab.

Inline Images

When Dr. Olivia generates an image (see Image Generation), it appears in-line in her reply bubble. A pulsing "CRT-scan" placeholder shows while the image is being created, then swaps in-place to the final image when it's ready. Click the image to open it full-size.

Rating a Response

Each of Dr. Olivia's replies has thumbs-up and thumbs-down buttons next to the timestamp. Click thumbs-up when a response was genuinely helpful, or thumbs-down when it missed the mark. Click the same thumb again to clear your rating. Ratings are private feedback to the development team — no score is visible to other students.

Continuing a Past Session

Past chats are listed under your conversation history. Selecting one drops you back into that session with full context. Generated images from previous sessions stay in the transcript as clickable thumbnails.

Ending a Session

The End Session button in the top-right of the header closes out your chat and takes you to a quick feedback screen, where you can rate the session and add an optional comment before returning to the avatar selection page. Clicking the PHSU logo does the same thing — it routes you through the feedback screen on the way back to selection. Either way you can skip the rating and go straight back if you prefer.

Tutor Status Indicators

While a reply is being generated, a small status line appears under the input:

Searching course materials… — pulling from your indexed lecture and textbook content.
Reviewing your USMLE/ExamSoft results… — fetching analytics.
Looking up Canvas assignments… — calling the Canvas API.
Looking up PubMed literature… — querying biomedical databases.
Generating image… — image-generation tool is running.

Same backend, different surface The text chatbot uses the exact same LLM pipeline as the avatar — same prompts, same tools, same materials. Switching between modes is a UI choice, not a different tutor.

Header Pills

The same Faculty / Acting-As pills described on the Selection page guide also appear in the chatbot header so you always know who Dr. Olivia thinks she's talking to.