Personalised Conversations
Every client gets a personalised copilot.
An AI copilot that answers your clients' questions using their real portfolio, goals, and risk profile. Scoped to your house views, held to your rules: it asks before it advises, and logs every word.
White-labelled and embeddable in your web and mobile apps.
A governed conversation
Nextvestment
Classified: Contextual · client context applied
Based on your moderate risk profile and 1-3 year horizon: FFLC is 50.8% of your portfolio and down 9.4% this quarter, above the concentration guideline for your profile. Worth reviewing: rebalancing toward your target allocation, or revisiting your horizon.
General information based on your profile. Not a personal recommendation.
Logged · advisor brief generated
Clients are already asking AI for financial advice.
Where clients turn first when a financial question arises
% of respondents
Responses reflect actions taken when a financial question first arises. Multiple selections were permitted.
Source: Nextvestment client survey.
It does not know them
Generic AI has never seen their portfolio.
It answers in the abstract, with no view of what they hold, what they are worth, or what they are trying to do. The advice is confident and unanchored.
It is not governed
No suitability. No disclosure. No record.
It will recommend without asking the questions you are required to ask, off your shelf, with nothing logged and no signal back to you.
How it works
Classified, scoped, answered, logged.
01
A client asks
In their own words, any hour, in your app. No script, no menu of canned questions.
02
It is classified and scoped
The question is read for intent and routed to the right analysis, with recommendations scoped to your universe.
03
It asks before it advises
If a recommendation needs more, it stops and asks for goal, horizon, and risk first. It will not guess.
04
Answered, and logged
A personalised, compliance-safe answer grounded in their portfolio, with a disclosure, every turn on the record.
It knows them
Every answer starts with their portfolio.
Before it answers, the copilot pulls the client's real context: their risk profile and horizon, their goals, their holdings and weights, the sectors they are exposed to, and what they have been researching. The answer is anchored to their portfolio, not to a generic investor.
The same picture your advisor would open with, on every turn.
A sample context
app.nextvestment.com/app/client
Jordan Lee
Investment profile
Risk tolerance
Moderate
Time horizon
1-3 years
Primary goal
Long-term growth
Experience
Intermediate
Portfolio
Recently researched
The guardrail
It asks before it advises.
Ask it to recommend something and, if it is missing what it needs to answer responsibly, it stops. It asks for the goal, the horizon, and the risk first, with structured choices, and holds the recommendation until it has them.
The suitability questions are enforced in the flow, not left to the model to remember.
A sample exchange
app.nextvestment.com
Recommendation request · suitability needed
Before I weigh in, a few details so the answer fits you:
What time horizon should I use?
How much risk are you comfortable taking?
Do you already own this security?
It will not make a recommendation until these are answered.
Made yours. Held to your rules.
The same AI your clients would reach for anyway, but answering inside your house: your universe, your sources, your suitability bar, with every word on the record.
Scoped to your house views
Answers stay inside the universe, sources, and house view you approve, with per-tenant grounding and an allowed set of news sources.
Asks before it advises
A recommendation will not proceed without the client’s goal, horizon, and risk. The copilot collects them first.
On the record
Every turn is logged, answers carry a disclosure, and outputs stay compliance-safe, no prescribing trades.
The proof
Live in a regulated market.
The AI copilot embedded in a regulated wealth platform in Singapore answers a self-directed client base under MAS regulation, on the shelf, on the record.
33,000+
users in the first year.
11%
convert to a trade each month.
3+ min
average session.
FAQ
Questions, answered.
What is it?
A governed AI copilot that answers your clients’ questions, personalised to their real portfolio, goals, and risk profile, scoped to your house and held to your rules.
Does it give financial advice?
It answers in context, and for a recommendation it asks for the client’s goal, horizon, and risk first and will not proceed without them. Answers carry a disclosure and stay compliance-safe; it does not prescribe trades.
How is it scoped to our house?
Answers stay inside the universe and sources you approve, with per-tenant grounding and an allowed set of news sources.
What makes the answers personalised?
Before answering, it pulls the client’s real context: risk profile, horizon, goals, holdings and weights, sector exposure, and what they have been researching. The answer is anchored to their portfolio, not a generic investor.
Is every conversation recorded?
Yes. Every turn is logged, and each release is run against scenario evals before it ships, so behaviour stays where you set it.
Is it in production?
Yes. It powers the AI copilot inside a regulated wealth platform in Singapore, under MAS.
Part of the Nextvestment platform. Recognised by
Make the answer yours.
Your clients will ask AI either way. Book a walkthrough and see the conversation answered in your house, on your terms.