An AI journal that asks follow-up questions
By The Architect team · Last updated
We make Architect. Everything this page says about follow-ups was checked against the app itself on October 23, 2026. The exchange below is a made-up example.
Yes. In Architect, a mentor reads your whole entry, beside a dated selection of what you wrote before, and answers it; a reply often ends with the question you were circling. You answer in a follow-up, and the conversation keeps going for as long as you want.
Unlike a printed guided journal, the question is written for this entry, not for everyone. Each reply to an entry or a follow-up counts as one reply on the free plan.
How follow-up questions work in Architect
- You write or speak an entry. Anything from a line to a long page, up to 25,000 characters.
- The mentor answers the whole entry. It reads what you actually wrote, with your earlier entries and their dates in view, and the reply can end with one pointed question.
- You answer in a follow-up. Type or speak your answer, up to 10,000 characters, or push back if the question missed. The same mentor replies, with the whole conversation in view.
- Keep going, or switch. Follow up again, or change mentor in the middle of the conversation for a different voice.
What a follow-up looks like
A short, made-up example. Nothing in it comes from a real person's journal.
The entry: "Got offered the team lead role. Everyone says it's obviously a yes. I said I'd answer Friday and I've been weirdly flat all week."
The mentor's question: "You wrote 'everyone says'. What do you say, when nobody is listening? Flat about the job, or flat about having to want it?"
Your follow-up: "Honestly, I like the work I do now. Leading means I'd stop doing it."
The mentor: "Then Friday's question isn't yes or no to a promotion. It's whether you're willing to trade the work you like for the title. What would you need to keep, to say yes?"
The question works because it could only have been asked of this entry: it picks up the writer's own words ("everyone says") and the detail they almost skipped (feeling flat). That is the test to run on any AI journal that asks questions.
A follow-up or another lens?
Architect has two ways to go further with an entry, and they do different things.
| A follow-up | See it through another lens | |
|---|---|---|
| What it is | Your answer, or your push-back, to the reply you got | A second opinion: one of the other mentors answers the same entry from its own angle |
| Who replies | The same mentor, with the conversation so far in view | A different mentor; Continue with that mentor carries the conversation on with it |
| How often | As many as you like | One per entry, right after a reply |
| How it counts | One reply each | One reply |
Use a follow-up to go deeper on what the mentor asked. Use another lens when you suspect the first mentor's whole reading of the entry is the one you would have given yourself. Both are on every plan. The four mentors.
How follow-ups count against your plan
Writing is free, and every entry saves. The free plan includes 11 full mentor replies, then one reply every 7 days. No card is needed. A reply to an entry and a reply to a follow-up count the same, one each, and a second opinion counts as one. In the Sitting, your spoken follow-ups count the same way.
The paid plan is $33 a month, $299 a year, or $9.99 a week, with no daily cap on replies to entries and follow-ups. Every number is on Architect facts.
Honest limits
- Not every reply asks a question. Sometimes the mentor names what it sees and stops there; you can still follow up.
- It can ask the wrong question. Say so in a follow-up; you can correct it in the conversation.
- It is not therapy. It is a tool for personal reflection, not a substitute for professional care.