Public Sector Canada

Digital Learning Strategy AI Assistant

The Challenge: Educators struggle to align course content with evolving digital learning standards, lacking tools to efficiently evaluate accessibility, inclusion, and flexibility across their materials.

The Solution: A GenAI-powered assistant that analyzes course outlines and delivers actionable feedback aligned with BC’s Digital Learning Strategy, supporting instructors, students, and administrators through a single intelligent platform.

Built on AWS: Amazon Bedrock powers LLMs for semantic analysis and course evaluation, with AWS Lambda for serverless processing and DynamoDB for scalable data storage.

Business Impact:

  • Instructors receive targeted suggestions to enhance accessibility, inclusion, and flexibility in course materials
  • Students access a chat interface for course support, while administrators monitor engagement trends through an analytics dashboard

Advancing Digital Learning: AI-driven course alignment removes the guesswork from meeting digital learning standards, making high-quality, inclusive education scalable across BC’s post-secondary institutions and beyond.

Visit the blog to learn more and access the open-source code.

Predicting Hospital Admissions

The Challenge: Emergency Departments face critical resource allocation challenges, manual assessments can’t keep pace with patient flow, leading to ED boarding, longer wait times, and preventable medication errors.

The Solution: An AI-powered admissions forecasting tool that predicts patient admission likelihood in real-time from triage data, enabling proactive resource planning across resourcing managers, pharmacy teams, and clinical staff.

Built on AWS: Amazon SageMaker with XGBoost and Random Forest models delivers real-time predictions, with Lambda for serverless processing, Aurora PostgreSQL for prediction storage, and AWS Amplify hosting the React dashboard

Business Impact:

  • Resourcing managers receive ranked, real-time admission scores (95.4% accuracy, 83.4% AUC) to proactively allocate staff and specialists
  • Pharmacy teams complete medication histories earlier in the admission process, preventing medication errors

Transforming Emergency Care: AI-driven admission forecasting shifts ED operations from reactive to proactive, reducing wait times and improving patient outcomes at scale.

Visit the blog to learn more and access the open-source code.

Research Data Insights Generator

The Challenge: Researchers spend weeks manually analyzing qualitative data, survey responses, audio files, to evaluate complex concepts like spatial empathy that traditional methods struggle to capture.

The Solution: An AI-powered prototype that automates qualitative analysis at scale, enabling researchers to upload data, define custom scoring criteria, and uncover patterns through LLM-as-a-Judge evaluation and RAG workflows.

Built on AWS: Amazon Bedrock powers multi-model ensemble scoring, enables context-driven RAG responses, Amazon Transcribe processes audio files, and Lambda with RDS PostgreSQL handles data processing and similarity search.

Business Impact:

  • Researchers transform weeks of manual qualitative analysis into automated, consistent insights across large datasets
  • Multi-model aggregation ensures reliable, reproducible findings

Unlocking Qualitative Research at Scale: AI-driven pattern recognition makes complex, multi-dimensional research concepts measurable, accelerating discovery across geography, social science, and beyond.

Visit the blog to learn more and access the open-source code.

Open Education AI Study Companion

The Challenge: Instructors lack the supplementary teaching materials commercial publishers provide, creating a critical barrier to open educational resource adoption.

The Solution: An AI-powered study companion that transforms open textbooks into interactive quizzes, flashcards, and conversational learning, giving instructors ready-made materials and students accessible, pedagogically-grounded support.

Built on AWS: Amazon Bedrock powers content creation through a RAG pipeline with multimodal embeddings, while AWS Glue processes large textbooks for seamless ingestion.

Business Impact:

  • Instructors generate materials from 400+ open textbooks, removing the #1 barrier to OER adoption
  • UDL-compliant features (audio narration, H5P exports) integrate with any LMS

Transforming Open Education: AI-generated practice materials close the gap between open resources and commercial quality, making inclusive education accessible at scale.

Visit the blog to learn more and access the open-source code.

Multi-Agent Framework for Assisting Quantum Materials Research

The Challenge: Quantum materials researchers face steep barriers to quantum computing adoption, slowing discovery and limiting access to quantum algorithms.

The Solution: An AI-powered multi-agent platform that automates quantum circuit generation from materials science data, enabling researchers to simulate and run quantum algorithms.

Built on AWS: Eight foundation models in Amazon Bedrock, orchestrate specialized agents via the AI Strands Agents SDK, with Amazon Braket connecting to quantum hardware through MCP servers.

Business Impact:

  • Researchers at all levels generate quantum circuits and access 150,000+ materials from the Materials Project database, no manual programming required
  • Multi-agent task decomposition reduces cognitive load and accelerates research workflows

Accelerating Quantum Discovery: AI-orchestrated agents combined with direct quantum hardware access put cutting-edge research capabilities in every researcher’s hands.

Visit the blog to learn more and access the open-source code.

Legal Aid Tool

The Challenge: Law students lack structured support to analyze complex cases independently, while supervising lawyers struggle to monitor progress and deliver timely feedback at scale.

The Solution: An AI-powered legal support tool that generates preliminary case summaries, guides students through interactive analysis, and gives supervising lawyers real-time visibility into student progress, without replacing critical thinking.

Built on AWS: Leveraging Amazon Bedrock for intelligent responses with AWS Lambda for serverless processing and RDS and DynamoDB for scalable data storage.

Business Impact:

  • Students receive AI-generated case starting points that deepen legal reasoning
  • Supervising lawyers monitor progress and deliver targeted feedback
  • Legal aid clinics expand capacity to serve low-income clients, addressing Canada’s access to justice gap

Closing the Justice Gap: AI-guided learning combined with human oversight transforms how the next generation of lawyers prepares for practice.

Visit blog to learn more and access the open-source code.

Document Smart Search

The Challenge: Government agencies and organizations managing vast document libraries face hours of manual searching to find relevant information among thousands of documents.

The Solution: An AI-powered solution that streamlines research and policy document exploration, integrating smart filtering with a conversational assistant to locate and summarize relevant information in seconds.

Built on AWS: Leveraging Amazon OpenSearch Service for hybrid keyword and semantic vector search, powered by Amazon Bedrock for embeddings and text generation, with serverless architecture using AWS Lambda, API Gateway, and AWS Glue for automated data ingestion.

Business Impact:

  • All Users locate relevant documents in seconds through conversational AI, advanced filtering, or analytics dashboards
  • Researchers & Policy Makers receive role-tailored AI responses with relevance scores and summaries
  • Organizations explore content without exact terminology, reducing search time from hours to seconds

Scalable Design: Supports any organization with large document collections, enabling efficient knowledge discovery across topics, mandates, and timeframes.

Ready to transform how your organization accesses institutional knowledge.

Visit blog to learn more and access open-source code.

AI Learning Assistant

The Challenge: Students struggle to get immediate help outside classroom hours, creating learning gaps and frustration.

The Solution: An AI-powered learning assistant delivering contextually relevant responses based on course materials, providing 24/7 student support with real-time progress tracking.

Built on AWS: Leveraging Amazon Bedrock for intelligent responses with AWS Lambda for serverless processing and RDS and DynamoDB for scalable data storage.

Business Impact:

  • Students receive instant, personalized support anytime, anywhere
  • Instructors monitor engagement through analytics dashboards showing concept mastery
  • Administrators manage courses and instructor assignments through a centralized portal

Visual Progress Tracking: Students see their learning journey shift from red to yellow to green as they master concepts, driving engagement and accountability.

Ready to transform student support and learning outcomes across your institution.

Visit blog to learn more and access open-source code.

Empathetic Communication Practice Tool

The Challenge: Pharmacy students lack safe, scalable opportunities to practice emotionally charged patient conversations outside the classroom.

The Solution: An AI-powered platform delivering real-time voice and text patient simulations with instant, personalized feedback across six empathy dimensions.

Built on AWS: Leveraging Amazon Bedrock with a multi-model AI architecture using serverless infrastructure for seamless scalability and real-time performance.

Business Impact:

  • Students practice anytime, anywhere with immediate coaching
  • Instructors gain analytics dashboards tracking progress and performance
  • Institutions integrate empathy metrics directly into curriculum and assessment

Ready to scale empathetic communication training across healthcare education.

Visit blog to learn more and access open-source code.