It does the work.
You approve it.
Built into your cap table from day one — it answers questions, extracts agreements, and drafts the work, then waits for your sign-off before a single number changes.
Most cap-table AI just chats.
Bolt a chatbot onto a database and you get confident summaries that can't be trusted with a single transaction — so you still do the real work by hand. An assistant that can't act safely isn't an assistant; it's a search box.
Ours is built in from inception — it proposes real actions against your live data, and never commits one without your approval.
A teammate for your cap table — not a chatbot.
Answers in plain language
Ask about ownership, dilution, or any transaction and get a precise answer drawn from your live cap table — not a generic guess.
Extracts your agreements
Upload a SAFE or convertible note and it pulls amount, cap, discount, and dates, then drafts the matching transaction for review.
Drafts investor updates
Turn this quarter's numbers into a board- or investor-ready update in your voice — you edit and send, it does the assembly.
Models scenarios on request
Ask it to price a round or refresh the pool and it runs the scenario against your live table, then explains the result.
Proposes, never mutates
Every change is a proposal with the exact edits shown. Nothing touches your records until you approve it — by design.
Knows what's missing
Ask "what's left for diligence?" and it checks the readiness list, then tells you the one thing standing between you and ready.
The answer, with the math behind it.
No exports, no formulas, no waiting on someone to "pull the numbers." Ask in plain language and get an answer you can act on — every figure traceable back to the live cap table.
- ✓ Natural-language questions across cap table and transactions
- ✓ Every figure traceable to its source record
- ✓ Follow-ups that remember the thread
- ✓ One step from an answer to a proposed action
It proposes. You approve. Nothing else.
Every action the assistant takes against your cap table is a proposal — shown in full, approved by a human, logged in the audit trail. The speed of AI with the control of a four-eyes review. No mutation without your sign-off, ever.
See it in your account →How the assistant works.
What can the AI assistant do?
It answers questions about your cap table and transactions in plain language, models what-if scenarios, extracts terms from SAFEs and convertible notes, and drafts investor updates. Anything that would change your records is proposed as a reviewable action, not executed automatically.
Does the AI change my cap table automatically?
No. The assistant proposes — you approve. Every action that would mutate your data is presented for review with the exact changes shown, and nothing is committed until you sign off. There are no surprise edits.
What AI model is it built on?
It's built on a leading frontier large language model, chosen for careful reasoning over your numbers rather than generic chat. The emphasis is on nuance and accuracy — not bullet-point summaries that sound confident and get the math wrong.
Can it read my SAFEs and convertible notes?
Yes. Upload an agreement and the assistant extracts the key terms — amount, valuation cap, discount, and dates — and drafts the corresponding transaction for you to review side by side with the source document before committing.
Is my data used to train AI models?
Your company's data is yours. It is used to answer your questions and do your work — not to train public models. Sensitive records stay within your account and your chosen storage.
The assistant works on everything here.
It does the work. You approve it.
Start free and put a cap-table teammate to work — one that never changes a number without your sign-off.