Twenty-plus years across engineering, data and finance, told as the seven questions I couldn’t leave alone. Walk the trail from the Atlantic to the Pacific, or skip to the summary.
Chapter 1 · 2000–2007 · Saint John, NB → Edmonton
What can a machine figure out on its own?
At UNB I wrote my honours thesis surveying quantum computer architectures and algorithms. At the University of Alberta, my MSc thesis used graphical models to predict protein function across the Gene Ontology’s hierarchy.
Side paths
- MSc thesis: Hierarchical prediction of protein function in the Gene Ontology using graphical models 2008
- Build a CNN for facial expression recognition 2018
Scroll west into town
Chapter 2 · 2008–2016 · Greater Vancouver
How does a business actually run?
Grad school gave me a strong interest in business, so I chose consulting to learn as many businesses as possible, fast. Numbers were the backbone of each. Inspired by my CFO, I spent two years of weekends earning my CPA.
- Senior Consultant / DeveloperIdeaca · 2008–2012
- Manager, Business Applications and IntelligenceZCL Composites · 2012–2016
Scroll on to the workshop
Chapter 3 · 2016–2018 · Greater Vancouver
Can I build it myself?
I co-founded Magic Beans to build 3D VR/AR showcases for complex products: packaging machines, aircraft interiors, home interiors. I led an offshore dev team, talked to hundreds of people at trade shows, and gave live demos onsite.
Side paths
- How much time will it take to cross the border at Peace Arch? 2018
- How many bikes will be shared in Vancouver next week? (part 1) 2018
- Why so many wildfires in BC lately? 2018
4 more posts from these years
Scroll on to the lighthouse
Chapter 4 · 2018–2023 · Pacific coast
If the data is big enough, will it tell the truth?
Not by itself. It takes the right questions, metrics and business context. At Visier, I cut customer onboarding from about six months to one. At Clearly, I moved on-prem systems to an AWS data platform for real-time insights.
Scroll to climb the lighthouse
Along this stretch of the trail
- AWS Certified Solutions Architect – Associate
- On-prem → AWS data platform
- Ad hoc report requests −30%
Side paths
- Organizational network analysis 2019
- Credit card fraud detection using SMOTE 2019
- Exploratory spatial data analysis: spatial autocorrelation 2021
10 more posts from these years
- People analytics: attrition predictions
- Things employees like and dislike about their companies
- 2019 HR Tech conference on Twitter, part 1
- 2019 HR Tech conference on Twitter, part 2
- Who are the top HR analytics influencers on Twitter?
- What people tweet about #coronavirus, in one minute
- COVID-19 Power BI dashboard
- Trump and Trudeau on Twitter during COVID-19, part 1
- Trump and Trudeau on Twitter during COVID-19, part 2
- Can you tell someone’s gender based on tweets?
Chapter 5 · 2023–now · Greater Vancouver
How do you lead a team to build better with AI?
I built a team from 0 to 10 engineers in 12 months, behind a grants management product used by 400+ universities and research institutions. We brought Cursor and Claude Code into everyday engineering work.
Scroll to the campfire
What the team built
- customer issue triage time, with a RAG-based triage assistant
- −40%
- incident backlog in 6 months
- −30%
I also launched and host an office-wide AI learning series.
The guide by the campfire will answer questions about my writing. Coming in the next build.
Side paths
- Build an LLM chatbot with PDF documents 2024
- Build knowledge-graph RAG with LlamaIndex from PDF documents 2024
2 more posts from these years
Chapter 6 · Aug 2024–now · Greater Vancouver · a side path
How do you serve your community with AI?
Since August 2024 I’ve served on the board of Connections Community Services Society, bringing practical AI to a community board. I also give AI talks in local schools, and build games with my kids.
Chapter 7 · Next · The trail continues
What’s the question now?
I’m exploring new questions now. Stay tuned.