The Problem
Every advisor who joins brings a book of clients. The book arrives in whatever format the advisor''s prior life produced. A clean CSV from Schwab, a screenshot of a TD spreadsheet, a PDF from a discontinued custodian, a printed list, an email body, a Word document, a Quicken export, a manually maintained Google Sheet that has grown over fifteen years. Account types are inconsistent (IRA, Roth, joint, trust, 529, all labelled differently). Fee schedules are inconsistent. Household groupings are missing. Beneficiary designations are scattered. Client identifiers are sometimes SSNs and sometimes nicknames.
The firm''s head of operations, head of trading, and executive assistant spend roughly 30 to 60 hours per advisor cleaning the book, deduping, account-typing, fee-categorizing, beneficiary-confirming, and entering everything into the portfolio management system before ACAT can even start. The work is unstandardized because the input is unstandardized. Errors compound (an account type miscategorized at intake misroutes fees for years).
This is the silent hidden cost of every new advisor relationship. Three operators lose a week to the next advisor''s spreadsheet, and the next one after that.
Multi-Format Parser
AI AgentIngests the advisor's book in whatever format it arrives in and normalises the structure
What The AI Does
Accepts CSV, Excel, PDF, screenshot, scanned image, email body, and Word document
OCRs paper inputs and image-based PDFs into structured data
Detects and parses every common custodian export format (Schwab, Fidelity, TD legacy, LPL, Pershing)
Captures every account, household, beneficiary, and fee schedule reference present in the source
Account Type & Fee Normaliser
AI AgentNormalises account types and fee schedules against the firm taxonomy
What The AI Does
Maps inconsistent account labels (IRA, Roth, Joint, Trust, 529, UTMA) to the firm's canonical taxonomy
Normalises fee schedules and billing arrangements against the firm's fee grid
Flags edge cases (legacy fee arrangements, transitional billing) for operator review
Catches errors at intake before they propagate into years of misbilled fees
Household Grouping & Beneficiary Detector
AI AgentGroups accounts into households, flags missing beneficiary data, and deduplicates
What The AI Does
Groups accounts into households based on relationship, address, and tax ID signals
Surfaces missing beneficiary designations against required account types
Deduplicates across accounts where the same client appears under different naming conventions
Produces the import-ready package for the portfolio management system
Head of Operations Import Approval
Human ReviewHead of operations reviews the standardised package, approves the import, and ACAT begins
Review Criteria
Expected Impact
Before:
Three senior operators block out a week to manually clean and import each new advisor's book before ACAT can begin.
After:
Any inbound book format becomes a standardised import-ready record. Operators spend hours, not weeks, and ACAT starts on day one.
Result:
70 to 85 percent reduction in roster intake hours per advisor, with zero account-type miscategorisation errors propagating into billing and reporting