The assistant answers questions about your real account data - balances, transactions, recurring activity, pay statements, and the patterns inside them. The clearer the question, the better the answer.
Types of questions
Spending
- “How much did I spend last month?”
- “What are my top spending categories?”
- “How much did I spend on dining out this week?”
- “Show me my Amazon purchases this year”
Income
- “What was my income last month?”
- “Show me my pay history”
- “How does my income this month compare to last month?”
Accounts and net worth
- “What’s my checking account balance?”
- “What’s my total net worth?”
- “Show me my credit card balances”
Trends
- “Is my spending going up or down?”
- “How has my net worth changed this year?”
- “What’s my average monthly spending?”
Specific lookups
- “Find my last payment to Netflix”
- “Show me all transactions over $100 this month”
- “When was my last paycheck?”
Cash flow and savings rate
- “Is my Cash Flow positive this month?”
- “What’s my savings rate?”
- “How much breathing room do I have?”
Starter prompts on data pages
If you’re new to chat, you don’t have to come up with a question from scratch. The Spending, Income, and Cash Flow pages each show a row of Ask AI prompt cards near the top - tap one to open the assistant with that question pre-filled. Useful when you want to see how the assistant phrases things and what kinds of answers it produces.
Quick tips
- Include a time frame when it matters - “last month”, “in February”, “over the last 6 months”
- Name merchants and categories explicitly when you have one in mind - “How much at Whole Foods?” beats “How much at the grocery store?”
- Use natural language - ask the way you’d ask a friend, not in keywords
- Follow up in the same thread to dig into anything that catches your eye - the assistant keeps the conversation’s context across messages until you start a new thread
Start broad with something like “How am I doing this month?” and follow up with specifics. The assistant is usually faster than you’d expect at narrowing down to the answer you actually want.