Corey Yang-Smith

Corey Yang-Smith

he/him

AI Team Lead @ Urban Systems | PhD Student @ University of Calgary

Hello, I’m Corey!

👨‍💻 Corey Yang-Smith

🌍 Based in Calgary, Alberta, Canada

Connect With Me

🚀 About Me

I’m an AI Team Lead at Urban Systems Ltd., where I specialize in bridging cutting-edge artificial intelligence with the real-world needs of civil engineering, construction, and public engagement. I turn emerging research into practical tools that drive measurable business impact—whether that’s streamlining operations, accelerating design, or enhancing decision-making.

🔍 Currently, I’m completing my PhD in Software Engineering at the University of Calgary, where I focus on automating software development workflows using LLMs. I’m deeply passionate about AI, especially in the areas of LLMs, generative AI, and multi-agent systems.

🏆 Career Highlights

🧠 AI Team Lead @ Urban Systems Ltd. (2023–Present)

  • 🚀 Built the AI Practice from Scratch: Spearheaded and scaled a company-wide AI initiative, scoping 100+ projects across 15+ business domains, unlocking over $10M in potential annual savings through intelligent automation and workflow optimization.

  • 👥 Team Growth & Leadership: Expanded the AI team from 1 to 4 members, overseeing hiring, onboarding, and strategic direction, while also leading technical execution and mentorship.

  • 📣 AI Education & Change Management: Delivered company-wide presentations on the use of AI tools and Microsoft Copilot to 300+ employees, playing a key role in driving technology adoption, change management, and digital transformation across the organization.

  • 🖥️ Infrastructure & Deployment: Designed and procured specialized AI/ML hardware, enabling secure in-house training and inference. Deployed large language model tools using Ollama, LangChain, Pydantic AI, LangGraph, Microsoft Copilot and DSPy.

  • 🎯 Smart Vision Systems: Developed a YOLOv8-powered object detection model for analyzing underground utility CCTV footage—cutting processing costs by 95% and dramatically speeding up defect identification.

  • 🧩 Graph Intelligence for Planning: Created a powerful graph-based ERP data visualization and forecasting tool using Neo4j + GraphSAGE, enabling smarter workforce planning. Integrated Redis for ultra-fast caching, Microsoft Entra ID + MSAL for secure access control, and added semantic search for intelligent data retrieval.

🧑‍🏫 Graduate Teaching Assistant (2024)

  • Facilitated Full Stack Web Development lab to over 40 students at the University of Calgary, emphasizing React, Node, and deployment best practices.

💬 Community Engagement & Talks

I love giving back! I’ve:

  • 📊 Apr 2025: Supplied a dataset and participated as an industry sponsor at the Data Science & Machine Learning Club Final Symposium.

  • 🏙️ Mar 2025: Organized the Urban Systems Case Competition with the University of Calgary, focusing on AI in Public Engagement, receiving over 70 entrants and 15 participating teams.

  • 🛠️ Feb 2025: Delivered a hands-on workshop titled “Introduction to Large Language Models” to 30+ students, guiding participants through real-world LLM tools and applications.

  • 🏘️ Jan 2025: Sponsored and attended the 2025 Westman Centre for Real Estate Case Competition, focused on Affordable Housing.

  • 🧑‍🏫 Jan 2025: Delivered a guest lecture to 15 students on “Introduction to Large Language Models” for graduate students at the University of Calgary.

  • 👨‍🔬 Jan – Apr 2025: Acted as an industry advisor for two graduate-level capstone projects, mentoring on graph machine learning and full-stack web development.

  • 🎓 Sept 2024 – Apr 2025: Acted as an industry advisor for two undergraduate-level capstone projects, mentoring on large language model development and full-stack web development.

🧪 Research & Exploration

🧵 AI for Software Engineering

🏗️ AI in Civil Engineering

  • Explored AI for flood forecasting, project planning, transportation engineering, land use optimization, domain-model (BIM) generation.

🎨 Generative AI & Diffusion Models

  • Tinker with ComfyUI workflows for diffusion-based image generation: restoration, colorization, video generation, and more.
  • Constantly experimenting with model architecture, fine-tuning, and optimization.

🛠️ Tech Toolbox

  • AI/ML: PyTorch, TensorFlow, Keras, LangChain, PydanticAI, LangGraph, transformers, GraphSAGE, YOLO
  • Dev Stack: React, FastAPI, Django, Node.js, Docker, Neo4j
  • Infrastructure: Ollama, vLLM, Redis, Qdrant, Elasticsearch, Open WebUI
  • Data: pandas, NumPy, Spark, Power BI, Tableau, FME