Clients now want portfolios built around them, not off-the-shelf funds. But personalising a portfolio and writing a bespoke quarterly report for each one was hand work, so an adviser could carry only a few dozen clients, the cost of serving the mass-affluent tier soared, and robo-advisers were taking the ground.
Tailoring an individual portfolio to each client, dropping a sector they didn't want, harvesting losses to offset tax, was manual work that simply could not be done across thousands of accounts. Personalisation was a privilege reserved for the largest clients.
Reporting was the other time sink. At every quarter-end, advisers spent weeks writing the letters that explain to each client why their portfolio moved, cutting and pasting charts and professional commentary one client at a time.
Buried in operations and paperwork, advisers had no time left for the business itself, no room to meet new clients or offer the services an existing client was ready for. The cost to serve climbed while the growth stalled.
Automate the portfolio work and the reporting, and give the adviser the judgement calls.
We wired an optimisation engine into the trading systems that rebalances thousands of portfolios simultaneously, respecting each client's own tax constraints, values-based preferences and exposure limits, and generating the trade orders automatically.
The engine builds each client's individual tax position into every rebalance, selling losers to offset gains where it helps, the active tax management that used to be impossible to run at scale.
A generative system takes the client's raw returns, sets them against the firm's market view, and writes a plain-language letter explaining how their portfolio performed and why, one per client, in minutes.
Before a client meeting, the copilot builds the relationship manager a full profile: upcoming tax events, portfolio analysis and suggested talking points ("client X is holding excess cash, propose a bond fund").
Every rebalance and every client letter goes to the relationship manager for a fast review and approval before it reaches the client, so the model does the work and the human owns the advice.
With the portfolio work and the reporting off their desks, advisers spend their time with clients and on new business instead of on operations, the part of the job that actually needs a person.
The service the billionaires got, now profitable for everyone.
The number of clients a single adviser could serve at the same quality rose 300%, which turned the mass-affluent tier from a loss leader into a genuinely profitable segment.
Producing personalised client reports fell from weeks of work to a single click with a fast adviser sign-off, and full personalisation with active tax management became a service the firm could offer every client, not just the largest.
Bespoke wealth service doesn't scale while a person hand-builds every portfolio and writes every report. The value here is an engine that personalises and reports across thousands of accounts within each client's real constraints, with the adviser signing off the advice, so quality holds as volume grows. It is the same generate-then-human-approve discipline behind our automated client-reporting work, aimed at portfolios and investors rather than the monthly accounts.
An accounting firm assembled every client's monthly report by hand from email and WhatsApp intake. AI built each report from the scattered sources with a human sign-off, turning days of assembly into minutes across dozens of clients.
Read the case → Customer success · SaaSChurn was decided months before renewal. A client-health model flagged at-risk accounts six months out and spotted expansion signals, lifting the renewal win rate to 68.5% and cutting gross churn 30%.
Read the case →
Reef TRH