Four habits that help
If you are a program manager
If you are a mentor
If you are a mentee
Keep going after the first answer
When an answer looks wrong

Iris understands ordinary sentences, so there is nothing to learn before you start. These are the habits that get you a better answer faster. For what Iris is and where to find her, see Meet Iris, your Mentorly assistant.

Four habits that help:

Ask for the thing you actually want. "Which mentors have not yet connected their calendar?" gets you a list of names. "Tell me about calendars" gets you a paragraph. Iris is better at a specific question than a broad topic.

Ask from the page you are on. Iris knows which page you have open, so you can say "these" and "this one". Open your members list and ask "how many of these have not completed onboarding?" and she understands the context.

Say the number you want. "How many" gets a count. "Which" or "list" gets names. "Break it down by" gets a table. If you got a paragraph and wanted a table, just say so.

Ask one question at a time. Three questions in one message get one blended answer. Three messages get three clear ones, and Iris remembers the thread so the follow-ups stay short.

If you are a program manager:

You have the richest data, so this is where Iris saves the most time. Things worth asking:

"How many mentees are not yet matched?"
"Which mentors are over capacity?"
"How many mentors have not set up any availability?"
"How many sessions happened this month?"
"How many mentees have completed at least one goal?"
"How many members have not completed onboarding?"
"What are the top skills mentees are looking to acquire?"

She also answers product questions, so you can stop hunting through the Help Center mid-task: "How do I add participants?", "How do I send a message to all Mentors at once?", "How do I upload Resources?"

And she draws on the mentorship guides, which is the part people miss. "How should first-time mentees prepare?" gives you something you can send to your cohort.

If you are a mentor:

"Can you tell me about my mentee and their progress?"
"Can you summarize my thoughts on past sessions?"
"What is the objective of this program?"
"How do I connect my calendar to Mentorly?"
"Can I use Zoom for my mentoring sessions?"

If you are a mentee:

"When is my mentor available for our next session?"
"Have I made any progress since I started being mentored?"
"How do I prepare for my first mentorship session?"
"Will my mentor be notified when I book a session with them?"
"When does this program end?"
"What is the objective of this program?"

Keep going after the first answer:

The first answer is rarely the end of it. Iris keeps the thread, so follow-ups can be short:

"Now only the ones in Montreal."
"Put that in a table."
"Which of those have a session booked?"
"Why does that number look low?"

She often offers the next step herself, such as formatting a list as a table or splitting it by mentors and mentees. Taking her up on it is usually faster than retyping the question.

When an answer looks wrong:

Iris is new, and she can be wrong. If something does not match what you expected:

Check the source. Answers about the product carry a link to the Help Center article behind them. One click tells you whether the article is wrong or Iris misread it.
Ask her to show her work. "Where did that number come from?" is a fair question, and a useful one.
Rephrase once. A question that names the thing directly often lands where a vaguer one did not.
Tell us. Click Message the Mentorly team in the panel. A message box opens with your last question already in it. What you send goes to the team along with the conversation it came from, so they can see what Iris actually said, and they reply by email. Wrong answers are the most useful thing you can send us right now: it is how Iris gets better.

Nothing Iris does can damage your program. She reads, she cannot change anything, and she cannot see past your own permissions. See What Iris can and cannot see for the detail.