Guided chatbot fallbacks for 760K+ BMO users
Correct responses: 45% → 86.5% in the first week
Read case study
Correct responses: 45% → 86.5% in the first week
Read case study629 entries audited and tree-tested
IA validated before launch, then migrated to Confluence: 112% month-over-month view growth and 72% engaged reads.
Read case study
18,000+ messages sent in 2 days
Built the business case, nearly doubled the character limit and shipped before the busiest week of the year.
Read case studyAlso worth a read: Style guides, and the environments that make or break them
10 people, 1 workshop, 3 solutions ready to ship
Read case study
Every team's error messages finally matched
A shared model for how errors explain what happened and how to recover.
Read case study
22% completion rate
Contextual guidance placed where support tickets showed people got stuck.
Read case study
Lowest ticket spike of any price increase
Across 40,000+ customers.
Read case study
100+ completions, every player ends on an instrument
Tested with real players and featured at Toronto Game Expo and Toronto Game Week. Case study coming soon.
Visit pageCurious how these come together? How the work gets done →
The titles have changed a few times. The work keeps landing in the same place: untangling messy information so people can find what they need, and so teams can keep it that way long after launch.
Full history on LinkedIn →
{{ item.a }}