The Chat: ask anything about anyone
The interface the whole product is built around. Type a question about any person you've ever logged or pulled from a connected source — and the AI answers in seconds, with the source attached.
/images/logo-dark.png
Ask, don't search
Search requires you to know the right keywords, the right date range, the right tool where the thing lives. Most of the time you don't know any of that. Ask is a different model. You describe the thing you want to know and the system brings it back. No keyword guessing, no tab switching, no trying to remember which tool you saved it in.
- No keyword guessing
- No tab switching across tools
- Describe, don't remember
What it actually feels like
Like having a chief of staff who's been paying attention to your relationships for months. One who never forgets, never gets tired, never loses context. Ask a question, get an answer with the source attached. Over time you stop trusting your own memory because you don't have to.
- Feels like a chief of staff
- Never forgets, never gets tired
- Builds the trust of an outside memory layer
Chat interface questions
What people ask the first time they use the chat.
What kind of questions can I ask?
Anything you'd ask a chief of staff. 'What did Sarah say her Q2 budget was?' 'What's open with Nick?' 'Who mentioned the partnership deal last month?' 'Pull a brief on my 10am meeting.' Plain English, plain-English answers.
- Anything a chief of staff would know
- Plain English, plain English back
Does it remember previous conversations with me?
Yes. The chat remembers context across sessions. You can pick up a thread from yesterday without re-explaining. The AI knows who you're talking about because you talked about them yesterday too.
- Persistent cross-session memory
- Pick up threads from yesterday
What if the AI doesn't know the answer?
It says so. The chat responds with 'I don't have data on that' instead of guessing. You always know whether the answer is grounded in your data or absent. Trust is built by refusing to hallucinate.
- Says 'I don't know' when it doesn't
- Zero hallucinated answers
Can I use the chat mid-meeting?
Yes — keep it open on a second screen. Ask live questions as the conversation develops. 'What did they push back on last quarter?' comes back in seconds with the source. Built for live use, not just pre-meeting prep.
- Built for live use
- Second-screen during meetings
Can I log new information through the chat?
Yes. Type a thought about a person — 'just had a call with Marcus, he's anxious about the launch' — and the AI extracts the entity, the context, and adds it to Marcus's journal entry. No forms, no fields.
- Log through chat, not forms
- AI extracts context automatically
Which AI model powers the chat?
The chat is powered by current-generation models with retrieval-augmented grounding against your journal. The model swaps as better options ship — the contract is that the answers stay grounded in your data, not the specific model version.
- Retrieval-augmented against your journal
- Model upgrades automatically as they ship
Get on the waitlist and lock in beta pricing.
Join the waitlistOpen the chat and start asking questions.
Free tier opens an empty chat. Log a few people and conversations, then ask questions about them the next day. That's the moment most operators convert.