I turned a cash process spread across four systems and a stack of spreadsheets into one clear flow advisors can move through without a manual.

I shadowed advisors through a real monthly close to find where the trust broke

I spent time with advisors during their actual cash-reallocation work, watching the month-end process step by step instead of asking them to describe it. Every month they pulled balances from Addepar, Salesforce, BNY-SEI, and KeyBank into a master workbook, then layered in private-equity calls, taxes, fees, and withdrawals by hand. The numbers were right by the end, but it took days, it broke whenever a linked formula broke, and one mistyped figure could carry through the whole month.

The bigger gap was forward visibility. The workbook only showed today. It never projected the running balance across pending transactions, so a coming shortfall surfaced only when someone worked the numbers manually. What advisors actually wanted was the earliest day each account would run short, early enough to fund it on purpose. That insight set the whole direction: the hard part was not a prettier dashboard, it was earning trust in a number that used to come from a spreadsheet they built themselves.

One workspace that consolidates the data, projects cash forward, and audits every move

I led design on a single platform that replaced the spreadsheet stack. It pulls portfolio, client, and balance data into one source of truth, automating the feeds that can be automated and standardizing the rest, then refreshes daily. On top of that, it turns today’s balance into a 60-day forward projection and flags the earliest negative-cash date across 150+ accounts, so advisors fund shortfalls before they happen instead of after.

Every cash movement runs as a validated, reversible, logged action. The platform checks fundability both ways before a trade is submitted, nets inter-account transfers to zero, and records every movement, approval, and document in an audit trail searchable by investor, account, and date for SEC review. Items stay out of the forecast until they are validated, so the projection stays conservative on purpose, and every figure traces back to the transaction that caused it. Monthly cash processing dropped from days to hours.

In a financial product, earning trust in the number was the real design work

Advisors had spent years building these numbers by hand and were not going to hand that judgment to a black box. So the work was less about the surface and more about making every projected figure inspectable, every validation visible, and every action reversible and logged. The dashboard was the easy part. The conservatism rules, the fundability checks, and the audit trail were what actually replaced the spreadsheet.