Financial Advisor

A personal financial support agent built using WhatsApp’s AI Studio and government sources to deliver accurate, up-to-date financial literacy guidance at scale.

00

problem

Newcomers to Canada face a maze of unfamiliar financial choices, made harder by language gaps, cultural differences, and biased advice.

solution

The opportunity lay in integrating the financial support agent into WhatsApp, allowing us to meet users on a platform they already trust while enabling rapid scalability and broader reach. The agent crawls government sources to provide accurate and current data to users in their native language if needed.

During an intensive 2 day hackathon, our team identified a challenge within the financial wellbeing space, defined a clear user group, and proposed a viable solution to a panel of industry experts. We placed Second!

The compressed format required us to move from problem definition through research, concept development, business viability, and final presentation within three days, with limited time for domain immersion or validation. To manage this, we intentionally spent 80% of our time on problem framing and exploration, and 20% on execution, ensuring the concept was defensible and grounded in real user needs rather than surface-level polish.

Strategic Rationale: Why This Population

We focused on newcomers to Canada based on three strategic considerations:

  1. Scale and impact
    Immigrants make up almost a quarter of the Canadian population. According to the 2021 census, 23% of people counted are or have been landed immigrants or permanent residents.


  2. Policy alignment
    Under the 2026–2028 immigration plan, the Government of Canada is reducing overall immigration levels while prioritizing economic immigration to fill labour market needs.


  3. Life-stage relevance
    Nearly two-thirds (64.2%) of recent immigrants fall within the core working-age group of 25 to 54, a period marked by frequent and high-impact financial decisions.



Research & Key Insights

Despite limited time, interviews with fellow hackathon participants revealed consistent patterns:

  • Localization, not literacy, is the core challenge
    Participants were generally confident managing finances but struggled with the Canadian-specific system; unfamiliar institutions, tax rules, and policies created hesitation and uncertainty.


  • Financial behaviour is largely reactive
    Many participants engaged with financial tasks only when deadlines arose, such as tax season.


  • Information overload undermines trust
    Financial information can change. Participants expressed concern about outdated or conflicting resources, which reduced confidence and delayed action.



Design Implications & Opportunity Areas

Research revealed an opportunity to meet users on trusted platforms and deliver on demand, up-to-date guidance by exclusively crawling Canadian government websites, ensuring information remains accurate and credible.

Together, these insights point to a service model focused on progressive, time-based support, high-trust information delivery, and emotional reassurance at high-stakes moments, scaled through a lightweight AI–human approach.


Solution Concept

We proposed an AI-powered financial support agent designed specifically for newcomers navigating the Canadian financial system. Rather than acting as a bank or generic financial literacy tool, the agent provides localized, current, and contextual guidance, adapting to the user’s stage of settlement and decision risk.

The agent crawls government sources to ensure information remains accurate and up to date, and can deliver guidance in a user’s native language when needed by reducing friction at moments of uncertainty.

The Features include 1) The individual decides how much or how little information to share. 2) Comparison examples to home country for examples. 3) External Links to bank products or more information.


Channel Strategy: Meeting People Where They Are

To maximize trust, adoption, and scalability, we integrated the financial support agent into WhatsApp. This allowed us to meet users on a platform they already use and trust, while enabling rapid global reach. Including support before arrival in Canada.

In the financial context, familiarity with the delivery platform plays a critical role in adoption. By leveraging an established communication channel, we reduced onboarding friction and transferred platform trust directly to the service.


Outcome & Reflection

Our team presented a complete, end-to-end solution spanning problem framing, research, service design, and business rationale, to a panel of experts and was awarded second place. Judges responded strongly to the clarity of the problem definition and the feasibility of the proposed approach.

This project reinforced the value of strategic restraint, deep problem framing, and cross-disciplinary collaboration under pressure. It also sharpened my interest in design strategy roles that operate at the intersection of policy, technology, and human behaviour, where design decisions can influence systems at scale.

year

2024

year

2024

year

2024

year

2024

timeframe

2 Days

timeframe

2 Days

timeframe

2 Days

timeframe

2 Days

tools

Miro, Figma

tools

Miro, Figma

tools

Miro, Figma

tools

Miro, Figma

category

Service Design

category

Service Design

category

Service Design

category

Service Design

01

The diagram shows a WhatsApp-based AI support agent where user messages flow through Twilio to a Rails backend that prompts an open-source LLM, stores user progress in a database, schedules automated tips, and feeds analytics for continuous improvement.
The diagram shows a WhatsApp-based AI support agent where user messages flow through Twilio to a Rails backend that prompts an open-source LLM, stores user progress in a database, schedules automated tips, and feeds analytics for continuous improvement.
The diagram shows a WhatsApp-based AI support agent where user messages flow through Twilio to a Rails backend that prompts an open-source LLM, stores user progress in a database, schedules automated tips, and feeds analytics for continuous improvement.
The diagram shows a WhatsApp-based AI support agent where user messages flow through Twilio to a Rails backend that prompts an open-source LLM, stores user progress in a database, schedules automated tips, and feeds analytics for continuous improvement.

02

A group of strangers come together to design something new over a weekend.
A group of strangers come together to design something new over a weekend.
A group of strangers come together to design something new over a weekend.
A group of strangers come together to design something new over a weekend.

.say hello

i'm open for freelance projects, feel free to email me to see how we can collaborate

.say hello

i'm open for freelance projects, feel free to email me to see how we can collaborate

.say hello

i'm open for freelance projects, feel free to email me to see how we can collaborate

.say hello

i'm open for freelance projects, feel free to email me to see how we can collaborate