Tech, ExplainedFintech
Part

Next best action for wealth advisers

What to offer each client next.

↑ Wealth management
INSIDE: WEALTH MANAGEMENT PLATFORMHOW IT WORKS · ₹40 LAKH LYING IN A SAVINGS ACCOUNTThe model reads theclient's holdings:₹40 lakh in savingsfor six monthsIt sees that a ₹25lakh fixed depositmatures next weekIt ranks threesuggestions, with adebt fund first, andshows the reasonThe adviser reads thereason and decides tocall the clientThe adviser recordswhether the clientaccepted, and themodel learns from itINSIDE: WEALTH MANAGEMENT PLATFORMHOW IT WORKS · ₹40 LAKH LYING IN A SAVINGS ACCOUNTThe model reads theclient's holdings:₹40 lakh in savingsfor six monthsIt sees that a ₹25 lakhfixed deposit maturesnext weekIt ranks threesuggestions, witha debt fund first,and shows the reasonThe adviser reads thereason and decides tocall the clientThe adviser recordswhether the clientaccepted, and the modellearns from itINSIDE: WEALTH MANAGEMENT PLATFORM · ₹40 LAKHLYING IN A SAVINGS ACCOUNTThe model reads the client's holdings:₹40 lakh in savings for six months↓It sees that a ₹25 lakh fixed depositmatures next week↓It ranks three suggestions, with a debtfund first, and shows the reason↓The adviser reads the reason and decidesto call the client↓The adviser records whether the clientaccepted, and the model learns from it
A model reads a client's idle ₹40 lakh and a maturing deposit, ranks three suggestions, and the adviser chooses one and records the outcome.
opens a page inside this one a page in the same scene a page in another scene a concept page a label only, with no page

Next best action for wealth advisers is an AI model that suggests which product an adviser or a banker should offer each client next, and why. It is also called a product recommendation engine. It runs inside the wealth management platform or the bank's customer software.

An adviser may look after a hundred or more clients. The adviser cannot study every client's account every day. The model can. It reads each client's holdings, transactions and life events, and it ranks a few suggestions for each client.

Take, for example, a client who has kept ₹40 lakh in a savings account for six months. Savings accounts pay low interest. The model also sees that a fixed deposit of ₹25 lakh matures next week. It suggests that the adviser call the client and discuss a debt fund. It also shows the reason, so that the adviser can explain it.

The adviser makes the final decision. The model only drafts the suggestion. This is an example of human in the loop: a model prepares the work, and a person checks and approves it.

This matters for two reasons. First, the suggestion must suit the client, not just earn a commission. In India, SEBI rules decide who may give investment advice and how. The page on investment advice regulation explains them. Second, the bank must be able to show why each product was offered. So the model records its reasons.

Banks use the same kind of model outside wealth management, for example to suggest a credit card or a loan to an account holder. An AI banking assistant can then make the offer directly to the customer in the app.

Concepts

Investment advice regulation

Who builds it

InvestCloud Inc.Advisor360°AlpheyaTandems
Tech, Explained is a part of ingrid.fyi. Every page explains one piece of technology in plain words.