top of page

VK

Image by UX Store

AI Chatbot Ontariocolleges.ca

Frame 36.png

Problem:

OntarioColleges.ca is the centralized platform for students to explore and apply to college programs across Ontario. However, OCAS was receiving a high volume of support tickets, emails, and phone calls from students struggling to find relevant programs that matched their interests or background.

These issues not only created frustration for students but also placed a heavy burden on our call center, who often had to guide users step-by-step through the platform.

Objective:

Improve the program discovery experience by designing a guided, AI-assisted solution that could help students find programs tailored to their needs—reducing dependency on customer support and improving user satisfaction.

Solution Overview:

We designed and launched an AI-powered assistant that guides students through a short series of questions, then intelligently recommends programs based on their responses.

The assistant simplifies the decision-making process and provides personalized recommendations with clear actions, relevant filters, and links to apply.

My Role:

Senior Product Designer

Duration:

4 Months

Image by Tirza van Dijk

IMPACT

  • 35% reduction in program-related support tickets within the first quarter

  • 25% increase in click-through rates to program details

  • +60% improvement in post-experience satisfaction scores

  • Recognized internally as a key step toward a more personalized and accessible student experience

My Role as Senior Product Designer

  • Identified the user pain points through analytics and stakeholder interviews

  • Co-created the solution vision with product and engineering

  • Led the design of the full experience—from the conversational flow to the result pages

  • Created user flows, wireframes, and high-fidelity mockups in Figma

  • Collaborated with AI engineers to align conversation logic with the design

  • Facilitated usability testing sessions and iterated based on feedback

  • Worked closely with the call center team to ensure the assistant handled top student queries effectively

  • Advocated for accessibility best practices across the flow

Key Features

Conversational Onboarding:

  • Guided Q&A flow:

    • What are your interests or career goals?

    • What's your education level?

    • Preferred learning mode: Online / In-person / Hybrid

    • Desired program duration

    • Budget constraints

  • Each question helps narrow down relevant programs dynamically.

Smart Filtering & Results:

  • AI returns a curated list of programs with:

    • Duration filters (e.g., 1-year, 2-year)

    • Learning mode (Online, Offline, Hybrid)

    • Tuition cost ranges

    • College name, location, and direct “Apply” links

    • Sorting options (by cost, popularity, deadlines)

Contextual Help:

  • Inline tooltips and definitions (e.g., explaining program types)

  • Option to restart or change answers at any point

  • Personalized list in PDF format is provided at the end

Frame 1.png
Frame 2.png
Frame 37.png
Frame 38.png

Research & Validation

  • Internal data audit showed high bounce rates on the program search page and a spike in support tickets near deadlines.

  • Call center interviews revealed students were confused by program naming, unclear filters, and information overload.

  • User testing showed the AI assistant reduced confusion and improved confidence in selecting programs.

  • A/B testing with the legacy search vs AI flow showed:

    • 40% higher engagement

    • Reduced average time to find a program from 6 mins to 2.5 mins

Key Learnings

  • AI-assisted UX must balance guidance with control - users should always feel in charge

  • Deep collaboration with non-design stakeholders (like the call center) led to more relevant and usable outcomes

  • Testing early assumptions with students saved weeks of rework and kept the solution grounded

bottom of page