2025•Fraser Health Authority

Calexis Chatbot

Simplifying IT request workflow and onboarding with an AI chatbot tool.

ChatbotAI LLMMachine Learning Tool

Team

Managing Consultant, Data Engineers, Data Analyst, Data Scientists

Duration

January - June

Role

UX/UI Designer

Overview

The Advance IT team at Fraser Health manages requests from staff regarding the Meditech Expanse platform, which is a tool that supports healthcare information systems and manages data.

The Problem

The Advance IT team at Fraser Health faces significant challenges due to the high volume and complexity of IT support requests. To find solutions, staff must manually search through numerous Quick Reference Guides that span hundreds of pages, consult scattered source materials across various support websites, or use a public-facing chatbot which offers only a narrow scope of support.

Solution

The client requested a new AI chatbot tool, called Calexis, to be a centralized internal platform that uses LLM-powered search to:

  • Aggregate QRGs, documentation, and Advance-related support material.
  • Provide user-friendly features that support improved response time.
  • Extend the scope of the legacy Advance Virtual Assistant (AVA) chatbot, but limit use to the internal IT team.

Projected Results

  • Free up Resources: by decreasing query resolution time by 50%, staff are able to address more queries per hour.
  • Increased User Satisfaction: An AI chatbot with pre-loaded QRGs and documents would reduce time spent searching for answers.
  • Scalable framework: Future tools can be developed to support other areas in the Fraser Health hospital and community network.

My Contribution

Discovery & Research

  • Conducted research on AI assistant interface best practices and features.
  • Stakeholder interviews with Advance IT managers and specialists to understand their workflow and pain points.
  • Audited Advance Virtual Assistant (AVA) and existing processes to uncover key usage gaps.

Design & Testing

  • Created wireframes and high-fidelity designs using Figma
  • Consulted with Advance IT team members for feedback and validation
  • Ran usability tests with Advance IT team members for further feedback and support

Key Takeaways

  • All-in-one platforms are intuitive and can significantly streamline staff workflows.
  • For AI chat tools to be effective, they must have user-friendly design and strong supporting features.
  • Close collaboration with IT and data teams was critical to building a solution that was both technically feasible and genuinely useful.

Next Project

View Next Project