November 2023 – March 2026
Student support at Sheridan College was fragmented across multiple websites, departments, and communication channels, making it difficult for students to find answers and creating inconsistencies in support responses. Starting in Comm100 and later expanding into ServiceNow, this project focused on bringing multiple departments into one system that unified knowledge, workflows, and support pathways.
Student Service teams were overwhelmed by increasing inquiry volumes while continuing to rely heavily on traditional email and phone support. At the same time, staffing resources were reduced due to external pressures such as budget constraints and declining international student enrollment.
Support requests were often routed inconsistently, with duplicate emails sent to multiple departments and large CC chains creating repetitive work. This increased wait times, reduced response efficiency, and made it difficult for students to navigate an ecosystem that did not align with their understanding of the institution.
Institutional language and department-specific terminology made information difficult for students to interpret. Resources were fragmented across systems, creating no clear source of truth and increasing dependency on frontline staff for clarification.
Students struggled to navigate institutional systems built around internal departments rather than student needs. Unfamiliar terminology, unclear ownership between teams, and disconnected support channels made finding the right information difficult, increasing confusion, repeated inquiries, and frustration.
Comm100 → ServiceNow
Comm100 → ServiceNow
The chatbot experience began in Comm100, helping Student Services move away from traditional email and phone support. However, ticketing remained disconnected from IT systems. Transitioning to ServiceNow created opportunities for stronger integration while requiring workflows to evolve alongside platform migration.
Early AI Limitations
Early AI Limitations
Generative responses were dependent on curated knowledge base content and platform capabilities available at the time. Early ServiceNow functionality relied on AI-assisted search to surface relevant knowledge articles, with fully contextual generative responses introduced as platform capabilities matured.
Balancing Immediate Needs
Balancing Immediate Needs
Growing inquiry volumes, staffing pressures, and changing institutional priorities required balancing immediate support improvements with longer-term conversational strategy. Solutions needed to reduce frontline strain while remaining flexible enough to adapt to evolving student and business needs.
Student Interviews
Understood needs, pain points, and expectations for support.
Chat Logs
Reviewed chat data to identify patterns, common questions, and drop-offs.
Competitive Review
Studied peer institutions to benchmark features and identify gaps.
Best Practices
Researched conversational UX patterns and chatbot guidelines in higher ed.
Research Insights
Synthesized insights around student needs, behaviors, and expectations.
Guided Flows
Created guided paths to help students find answers faster.
Generative Support
Integrated content and AI responses to provide relevant answers within the chat.
Human-in-the-Loop
Enabled easy escalation to live agents when students need personalized support.
Content Strategy
Prioritized high-impact topics for clearer answers.
Insight
Students often struggled to identify where to go for help and preferred guided support that reflected how they naturally searched for information.
Design Response
Created guided conversation flows and suggested questions to reduce ambiguity and help students find relevant answers faster.
Insight
Students often left the chat to search across multiple websites when answers felt incomplete or difficult to find.
Design Response
Integrated trusted knowledge and generative support directly into the experience to reduce unnecessary navigation and keep students in one support flow.
Insight
Some support needs involved exceptions, account-specific issues, or situations that required reassurance from staff.
Design Response
Designed clear escalation pathways to help students transition to human support when automation was no longer enough.
Insight
Students wanted faster ways to access support without navigating institutional language or complex systems.
Design Response
Used guided topic selection and simplified entry points to reduce effort and help students reach answers faster.
Insight
Support experiences varied across departments, creating inconsistent expectations and responses for students.
Design Response
Created shared conversational patterns, guided flows, and content structures to support a more consistent experience.
Insight
High-frequency questions repeatedly reached frontline teams, increasing response volume and slowing support.
Design Response
Prioritized guided flows and knowledge-driven responses for common topics to reduce repetitive inquiries and surface answers faster.
Guided flows supported common or complex topics that required structured language, clear next steps, and predictable pathways.
Greeting Menu
Personalized greeting based on student status, account access, or guest entry.
Topic Selection
Students select a topic and guided prompts clarify the support they need.
Guided Options
Suggested questions and branching paths guide students toward relevant support.
Deliver Support
Relevant answers are surfaced through curated content, guided support, or knowledge articles.
Solution Provided
Student receives the answer, resource, or next step.
Escalation Required
Issue escalates to front-line staff when additional support is needed.
Guided path example in chat
Generative support handled questions without predefined paths, allowing students to ask naturally while AI surfaced trusted answers from existing knowledge sources.
Generative flow example in chat
Open-ended Question
Students ask questions naturally using their own words.
AI Understands Intent
AI interprets intent to identify the student's support need.
Search Trusted Sources and Generate Response
AI searches trusted sources e.g. web pages, knowledge articles and other relevant information.
Deliver Generative Support
AI provides contextual answers with links to trusted resources.
Solution Provided
Student receives the answer, resource, or next step.
Escalation Required
Issue escalates to front-line staff when additional support is needed.
Structured prompts and guided pathways helped students navigate common support topics more confidently, reducing friction in high-volume and procedural requests.
Knowledge-driven responses prioritized institutional content, helping students find more reliable answers without needing to navigate multiple systems or webpages.
A dual-mode design gave students the right type of support for each situation — structured flows for common requests and generative responses for open-ended questions.
The conversational model created a more sustainable approach to student support by balancing self-service, contextual guidance, and escalation when human assistance was needed.
This project reinforced the importance of designing within constraints. Rather than replacing existing systems, the focus was on improving how students moved through them. Working within Comm100 and ServiceNow required thoughtful content structure, clear escalation paths, and support experiences that felt connected and scalable.
One key takeaway from this project was that AI should support people, not replace them. While conversational tools can help students find answers faster, some situations still require human support. An important part of the design was creating clear and natural pathways to staff when automation was no longer enough.
If this project continued, I would focus on creating more personalized support experiences based on student needs and context. I would also explore testing and analytics to better understand friction points, refine guided pathways, and strengthen how conversational and structured support work together.
