Building an AI Wealth Management Application for Singapore Financial Advisers

A team of Singapore financial advisers wanted to serve more clients without losing the personal advice that sets them apart. Thus, Tech JDI built a wealth management application in which an AI assistant answers everyday questions, collects each client's details, and maps their goals. 

Client Background

Our client, a team of experienced, licensed financial advisers in Singapore, set out to launch a wealth management application for working professionals and young families. The app would help users plan for key life milestones, from buying their first home or upgrading to a condominium to funding their children's education and preparing for retirement.

Their goal was to provide instant, personalised financial guidance while making it easy for users to connect with a human financial adviser whenever they needed more support.

Challenges

As enquiries grew, advisers found themselves spending more time on routine tasks while clients waited for the financial guidance they actually needed.

The team needed an app that could automate routine enquiries, streamline client onboarding and provide on-demand financial projections, while keeping all financial recommendations under the oversight of licensed advisers.

Our Solution: Customized AI Solutions for a Client-Facing Wealth Management Application

The client reached out to Tech JDI to help design and build the entire application, from the user experience and AI assistant to the financial planning engine powering it. 

The aim was to give advisers more time for conversations that require their expertise, not to replace their judgement with AI. The assistant handles explanations, data collection and financial projections, while licensed advisers remain responsible for recommendations.

Through the app, clients can get instant answers to common questions, share their financial details through a guided conversation, and see projections based on their goals. When they need professional advice, the app passes their information and a summary of their needs to an adviser, so the conversation can start with planning rather than paperwork.

The solution comprised five components, each designed around the client's original requirements. The first three, supported by core market data, formed the 12-week MVP, while the remaining components were delivered in Phase 2.

1. Gen AI Chatbot With Adviser Handover

The client-facing AI assistant runs on a large language model through Azure OpenAI Service, drawing answers exclusively from the team's approved materials, including product factsheets, fee schedules, FAQs and financial planning guides. If the information isn't available, it acknowledges the gap rather than guessing.

When a question crosses into personalised financial advice, such as choosing a fund, determining insurance coverage or deciding whether to switch investments, the assistant hands the conversation over to a licensed adviser. The adviser receives a summary of the conversation and the client's profile, allowing them to pick up where the AI left off without asking the client to repeat themselves.

To keep responses reliable and advice within the right boundaries, the app uses multiple layers of safeguards. A classifier and keyword rules flag advice-seeking questions, low-confidence responses are routed to an adviser, and every answer cites its source document. A Talk to an Adviser button also remains visible throughout the experience, giving clients a direct route to human support.

The safeguards are reinforced through weekly reviews, where advisers sample conversations to identify inaccurate answers and missed handovers. These findings are then used to refine the approved content and detection rules over time.

2. Singpass Myinfo and Guided Fact-Find

New users sign in through Singpass and, with their consent, use Myinfo to retrieve available government-held details, such as their name, NRIC, address and income records. These details automatically populate the sign-up form, reducing manual data entry. Myinfo streamlines information collection but does not replace the firm's customer due diligence processes, including screening and risk assessment.

Next, the AI assistant guides clients through a conversational fact-find, collecting details about their income, financial commitments, assets, dependants and goals one step at a time. It checks responses for missing information or inconsistencies, then organises the information into a structured client profile that the adviser can review before the first meeting.

3. AI-Driven Goal Visualisation

The app projects each goal based on the client's cash flow, savings rate, and assumptions about investment returns and inflation. For each goal, it runs 1,000 market scenarios to estimate the likelihood of success, such as a 95% chance of funding financial freedom by age 35.

The AI then translates these projections into detailed summaries, helping clients understand their results, such as taking on a larger home loan could affect their financial goals. 

It also displays the underlying assumptions, including expected returns, inflation, savings rate and time horizon, so clients can see how changing these inputs affects their results. As a result, advisers can now review this data before discussing a financial plan with the client.

4. Market Information Gathering

The app brings in market prices and economic data through licensed data feeds, along with property listings for home goals. A news summariser condenses market updates into short briefs for advisers each morning. When the data changes, for example, when interest rates move, projections are recalculated and approved rules flag the plans most affected for adviser review. 

A flag never triggers a recommendation or a client message on its own; the adviser decides what, if anything, to discuss.

5. Optimisation of Product Offerings (Phase 2)

The app analyses client questions by topic and tracks where users drop off during the fact-find process, with personal information removed and client consent obtained. A monthly insights dashboard highlights the most common questions, financial goals and gaps in the existing content. The team uses these insights to prioritise new products, guides and services based on clients' needs.

How We Build and Launch The App

The Results 

Users get instant answers to common questions, can check their financial projections anytime, and can speak to a licensed adviser when they need personalised advice. Since Phase 2, monthly insights also help the team understand what users need most and improve its products and content accordingly. 

AI in financial services isn't just about adding a chatbot. It's about reducing manual work, making services more accessible, and giving professionals more time to focus on decisions that require their expertise.

For this project, Tech JDI combined an AI assistant, automated onboarding and financial planning logic in one application, with clear boundaries around what AI could handle and when a licensed adviser needed to step in.

The same principle applies beyond wealth management: start with the workflow you want to improve, then build AI around the task it can do reliably.

Planning an AI-Powered Application for Your Business?

Whether you're launching a new digital product, adding AI to an existing platform or automating a workflow that still relies on manual work, Tech JDI can help you define the right solution and plan its development.

We'll work with you to identify where AI can add value, what needs to be integrated, and how to move from an idea to a production-ready application.

Frequently Asked Questions