The Assistant is an in-app AI helper that lives next to each client's plan. It's labeled Dora in the sidebar (Sparkles icon) — open it from inside any client and ask anything about that client. It can summarize the plan, surface gaps, draft client emails, look up tickers, import transactions from a bank statement, build a guide for the client, and propose changes to the data — all scoped to the client you're currently in.
It's still evolving, so the shape will keep changing.
Where to find it
Open any client and look in the left sidebar. Dora is the bottom item, below Guides. Only advisors see it — clients in the portal don't.
Each client has its own Dora. Threads belong to that client; switching clients switches threads.
What the empty state looks like
When you open Dora on a fresh thread you'll see a greeting ("Good morning, Taylor") and a one-line prompt: "What can I help you with for [client]?" Below the input box are a few suggested starter prompts to get you going. The exact prompts change as the product evolves, but they're shortcuts — anything that lives in the client's plan is fair game.
Things you can ask it:
Summarize a client's plan, open changes, tasks, goal funding, insurance coverage, and allocation vs. target
Prepare an agenda for an upcoming meeting — meeting prep requests come back with a visual meeting prep card, not just text
Update the plan profile (the high-level client facts Dora uses for context)
Pull up a ticker by name or symbol — Dora returns live price, fundamentals, sector mix, and a chart
Draft a client email or summarize the latest changes
Build a guide for this client (see Dora-built client guides below)
Import transactions from a bank or credit card statement (see Importing transactions below)
Visual answer cards
Dora doesn't just answer in text. For many questions it responds with a rich card rendered right in the chat — a visual summary built from the client's live data. Cards exist for things like:
Client snapshot — the household at a glance
Meeting prep — agenda, recent activity, and open items for an upcoming meeting
Next actions — what to do next for this client
Cash flow snapshot — income, expenses, and savings
Investment allocation — how the portfolio is invested
Goal progress — funding status across goals
Tasks and changes — open tasks and changes to be made
Insurance, estate plan, and debt summaries — coverage, document completeness, and liabilities
Tax cards — a tax scenario summary, planning opportunities, or a side-by-side scenario comparison
Most cards include action buttons that jump you to the relevant page in the app (the Goals page, a tax scenario, the task list) while the conversation stays open in a side drawer, so you don't lose your place in the chat.
How it works
Dora has read access to the full client store and can write JavaScript against a frozen snapshot of it to answer questions. When you ask it to change something — add an asset, update a goal assumption, mark a task complete — it doesn't write directly. It drafts a proposal, you see the diff in an approval panel, and nothing persists until you click approve.
That means:
Read-only questions ("what's their savings rate?", "show me top holdings") return an answer in the chat.
Change requests ("add a 401k for John worth 80k", "set the retirement goal to age 67") show up as a proposal with an approve/reject button. Reject and nothing happens.
Guide requests are the one exception: when you ask Dora to build a guide, it creates the guide immediately (no approve step) because a guide is a separate page you can review, edit, or delete on your own — it doesn't touch the client's underlying data.
You can attach files (paste an image, upload a PDF, drop a CSV) and Dora reads them as part of the question. If you attach a document without saying what to do with it, Dora asks first: extract the data into Kerdora, or just keep the file in the document vault. Nothing is extracted until you choose.
Dora-built client guides
Ask Dora to build a guide — "build a retirement-readiness guide for this client," "make a one-page net worth summary" — and it assembles one from the same components you'd use in the guide editor, populated with the client's live data.
A few things to know:
The guide is created immediately and Dora posts a card in the chat with an Open guide button. There's no approve/reject step — building a guide doesn't change any client data.
It's a client guide, owned by that client file. You'll find it in the guide dropdown at the top of the Guides page under the Created for client section, alongside your regular templates.
It's fully editable. Open it and adjust it like any other guide — add or remove components, change text, rearrange rows. Changes save to that client's copy only; your firm templates aren't touched.
If you don't like it, delete it. Nothing else in the plan is affected.
Importing transactions
Attaching a bank or credit card statement (CSV or PDF) and asking Dora to import the transactions kicks off a guided import flow. Dora figures out which account the statement belongs to, asks you to confirm if it's unsure, and produces a Transaction import card showing transaction count, date range, inflows, outflows, and net totals.
Approving the card runs the import as a background job with visible progress phases: queued → classifying → importing → complete (or failed / needs attention). The card updates as it runs so you can watch the imported/failed counts move. See Importing Transactions from a Bank or Credit Card Statement for the full walkthrough.
Threads
Each conversation is its own thread. You can have many per client. The URL updates as you switch — /client/<id>/agent/<thread-id> — so threads are bookmarkable and shareable inside your team.
Start a new thread by going back to Dora in the sidebar (the empty state with the suggested prompts is the "new thread" view).
Thread titles
Threads get short titles automatically. After your first message, Dora generates a concise title from the conversation so the thread list isn't a wall of "Untitled." You can rename any thread by hand; once you do, the manual title sticks and won't be overwritten by the auto-titler.
What it can't do
Reach across clients. Each Dora is scoped to the client you're in. It can't compare two clients or pull data from another file.
Touch anything outside the client store. No CRM, calendar, email send, or third-party API access.
Change client data without your approval. Every data write is a proposal you approve or reject. (Guides are the exception — they're created directly because they don't alter the plan's data.)
Auto-triggered runs after document extraction
When the Extractor finishes processing a batch of documents for a client, Dora kicks off a thread automatically to review what was found and flag anything worth your attention. You'll see the new thread under Dora for that client — open it to see the review and approve or ignore any proposed follow-ups.
If you'd rather skip the auto-review, just close it. It doesn't change anything until you act on a proposal.
Tips
Be specific. "What's their retirement gap assuming a 4% withdrawal rate?" beats "Are they on track?"
Approve carefully. The diff shows every event Dora wants to apply. Read it before clicking approve, especially for batch updates.
Use it for first drafts. Drafting a client email, summarizing the plan, or building a starter guide is where it shines. Treat the output as a starting point, not a finished deliverable.
Known limits
Long, multi-step changes can take a few rounds of back-and-forth to land cleanly.
It occasionally misses context that's only in a document until it loads that document explicitly. If an answer feels off, ask it to re-check using the document.
Very large transaction imports (8,000+ rows) can hit the agent's output-token limit during account/category setup. Splitting the file into smaller statements is the workaround.
Output formatting in the chat will keep evolving.
If something looks wrong or Dora proposes a change you didn't ask for, reject the proposal and message support — the trace helps us improve it.