Project Collection

Project Collection

Project Collection

Microsoft x BrainStation Hackathon

Microsoft x BrainStation Hackathon

Microsoft x BrainStation Hackathon

Overview 🔍

Overview 🔍

Working collaboratively with students from other disciplines, we tackled the challenge proposed by Microsoft: How might we support Microsoft customers adapt to AI innovations? Our solution was a specialized Copilot persona, designed to help small business owners craft better prompts and achieve more effective results with AI tools.

Overview 🔍

Working collaboratively with students from other disciplines, we tackled the challenge proposed by Microsoft: How might we support Microsoft customers adapt to AI innovations? Our solution was a specialized Copilot persona, designed to help small business owners craft better prompts and achieve more effective results with AI tools.

First Place Winner 🏆

First Place Winner 🏆

Our team won first place in the Microsoft x BrainStation 24-hour hackathon. The judges praised our innovative approach, highlighting the effectiveness of a plugin-based solution over designing an app or program.

First Place Winner 🏆

Our team won first place in the Microsoft x BrainStation 24-hour hackathon. The judges praised our innovative approach, highlighting the effectiveness of a plugin-based solution over designing an app or program.

Meet the Team 👥

Meet the Team 👥

UX Designer - Sean Lee

UX Designer - Anise Aleyasin

Software Engineer - Ching Hang Lam

Software Engineer - Tymur Rezaei

Data Scientist - Rodrigo Pelayo Ochoa

Data Scientist - Winston Yuen

Meet the Team 👥

UX Designer - Sean Lee

UX Designer - Anise Aleyasin

Software Engineer - Ching Hang Lam

Software Engineer - Tymur Rezaei

Data Scientist - Rodrigo Pelayo Ochoa

Data Scientist - Winston Yuen

My Role 👨‍💻

My Role 👨‍💻

As the lead UX designer on the team, I focused on designing intuitive interactions for the Business Consultant Copilot. I collaborated closely with cross-functional team members to create a seamless user experience, ensuring it aligns with Microsoft's existing design language

My Role 👨‍💻

As the lead UX designer on the team, I focused on designing intuitive interactions for the Business Consultant Copilot. I collaborated closely with cross-functional team members to create a seamless user experience, ensuring it aligns with Microsoft's existing design language

Prototype Highlight
Prototype Highlight
Project Impact
Project Impact
Problem Space
Problem Space
Analysis
Analysis
Design
Design
Reflection
Reflection
Back to Top ✦ Back to Top ✦

Prototype

Prototype

A specialized Co-pilot with guidance

A specialized Co-pilot with guidance

A specialized Co-pilot with guidance

Our proposed solution to increase trust in AI technology for Microsoft business customers is an specialized Co-pilot persona called Business Consultant. This persona would guide users in refining their prompt inputs, helping them achieve more accurate and effective results from Copilot.

Users are greeted with a few questions from Business Consultant before it generates a well-constructed "template" prompt for the user to put into the chat box and edit. Users has the freedom to type in the chat box at anytime.

Project Impact

Project Impact

🏆 1st Place Winner 🏆

🏆 1st Place Winner 🏆

🏆 1st Place Winner 🏆

Our solution secured first place in the competition, with the Microsoft judges recognizing the effectiveness of our plugin approach over a traditional app or program.

Problem Space

Problem Space

Low business usage in AI and low public trust in AI

Low business usage in AI and low public trust in AI

Low business usage in AI and low public trust in AI

Microsoft is committed to supporting Canada’s SMEs and enterprise businesses in their AI transformation journey. Microsoft aims to help drive responsible AI adoption, boost economic productivity and ultimately improve the quality of life for all Canadians.

With a background in banking, I chose to tackle a personal finance issue many Canadians face. What I discovered through my research:

With a background in banking, I chose to tackle a personal finance issue many Canadians face. What I discovered through my research:

However, only 9% of Canadian businesses are using Gen AI. At the same time, public trust in AI is low: only 31% of Canadians trust AI, 23% lower than the global average.

With a background in banking, I chose to tackle a personal finance issue many Canadians face. What I discovered through my research:

With a background in banking, I chose to tackle a personal finance issue many Canadians face. What I discovered through my research:

How might we support Microsoft customers adapt to AI innovations?

How might we support Microsoft customers adapt to AI innovations?

How might we support Microsoft customers adapt to AI innovations?

Research & Analysis

Research & Analysis

Here's what we found on Canadian SMEs:

Here's what we found on Canadian SMEs:

Here's what we found on Canadian SMEs:

We honed in on finding how many SMEs are in Canada and how many of them are using AI to support their business.

~ 135,000 SMEs

~ 135,000 SMEs

~ 135,000 SMEs

in Retail Trade

in Retail Trade

in Retail Trade

Second biggest industry in Canada

Second biggest industry in Canada

Second biggest industry in Canada

Only ~3.5%

Only ~3.5%

Only ~3.5%

of AI Usage in Retail

of AI Usage in Retail

of AI Usage in Retail

Second lowest usage in Canada

Second lowest usage in Canada

Second lowest usage in Canada

Trust = Prompt Engineering

Trust = Prompt Engineering

Trust = Prompt Engineering

Our take on the challenge was to guide people towards generating better outcomes by asking the right questions.

Here's how we believe it increases trust in AI:

  1. Provide clear and specific instructions

  1. Provide clear and specific instructions

  1. Provide clear and specific instructions

  1. Well-engineered prompts lead to consistent and reliable outputs

  1. Well-engineered prompts lead to consistent and reliable outputs

  1. Well-engineered prompts lead to consistent and reliable outputs

  1. Guides AI behaviours


  1. Guides AI behaviours


  1. Guides AI behaviours


  1. AI generates the right output, enhancing user trust and satisfaction with the AI system

  1. AI generates the right output, enhancing user trust and satisfaction with the AI system

  1. AI generates the right output, enhancing user trust and satisfaction with the AI system

But… Microsoft already have guides

But… Microsoft already have guides

But… Microsoft already have guides

We discovered that Microsoft already have documents and guides on how to create good prompts such as Copilot for Microsoft 365 Guide, Microsoft Learn and Copilot Lab.

Focused on retail business owners

Focused on retail business owners

Due to time constraint, we made some assumptions for small - medium retail business owners:

  1. They are busy individuals focused on running their business

  1. Many small - medium businesses don't have budgets to spend on enterprise level AI products

  1. Since they are busy individual focused on their business, they don't have time to go through Microsoft's prompt engineering documents

  1. Last of all, many of them just don't know how to use generative AI effectively

"Pedaling towards innovation, one ride at a time"

Mike Salivan

Mike Salivan

38 years old

38 years old

38 years old

Local bike shop owner

Local bike shop owner

Local bike shop owner

Retail Industry

Retail Industry

Retail Industry

Vancouver, Canada

Vancouver, Canada

Vancouver, Canada

"Pedaling towards innovation, one ride at a time"

"Pedaling towards innovation, one ride at a time"

Goals:

Goals:

Goals:

  • Improve inventory management

  • Improve inventory management

  • Improve inventory management

  • Enhance customer experience

  • Enhance customer experience

  • Enhance customer experience

  • Expand market reach

  • Expand market reach

  • Expand market reach

  • Offer specialized services

  • Offer specialized services

  • Offer specialized services

Pain Points:

Pain Points:

Pain Points:

  • Talent acquisition and retention

  • Talent acquisition and retention

  • Talent acquisition and retention

  • Market competition

  • Market competition

  • Market competition

  • Customer retention

  • Customer retention

  • Customer retention

  • Marketing limitation

  • Marketing limitation

  • Marketing limitation

Design

Design

We changed how prompts are tailor-made and displayed for the business user

We changed how prompts are tailor-made and displayed for the business user

We changed how prompts are tailor-made and displayed for the business user

In a nutshell:

  1. A specialized Co-pilot like Fitness Trainer, Vacation Planner and Designer called Business Consultant

  1. Users have selections on the screen to refine their area of focus

  1. At the end, Business Consultant will generate template prompts that users can edit

  1. The chat box remains open for any input from the user at anytime

Next Steps & Reflection

Next Steps & Reflection

What are some success indicators?

What are some success indicators?

Due to time constraints, we did not have time to test our prototype and validate our design, however, there are some success indicators that can help validate our idea:

  1. Adoption Rate:

  1. Adoption Rate:

The number of users that uses the new feature

  1. Active Users:

  1. Active Users:

Track the number of users actively using the Business Consultant

  1. Task Completion Time:

  1. Task Completion Time:

How long it takes for a user to enter a prompt with or without the feature

  1. Survey & Feedback:

  1. Survey & Feedback:

We can ask about the quality of the prompt (i.e like/dislike rating)

Thanks for checking this out!

Thanks for checking this out!

Like what you see? or want to say "hi"? Get in touch with me!

Let's Connect on LinkedIn!

Let's Connect on LinkedIn!

Let's Connect on LinkedIn!