01. The Context
At Moneris, DesignOps had become a bottleneck — not because of talent, but because of fragmentation. Design, product, and engineering were operating across disconnected systems, with too much manual coordination and not enough shared intelligence.
I led the effort to rethink this as an operating system problem, not a tooling problem.
02. Execution
We designed and implemented an enterprise AI interface layer, built on Open WebUI and a set of custom-tuned LLMs, to unify how teams access knowledge, generate work, and move from idea to execution. The goal wasn’t just efficiency — it was to create a consistent, scalable way for teams to collaborate and make decisions.
Four purpose-built agents were embedded directly into existing workflows:
- AI Interface Layer — Built on Open WebUI and custom-tuned LLMs to unify how teams access knowledge and generate work.
- Purpose-Built Agents — A scalable system of dedicated agents embedded into existing workflows to augment and automate specialized tasks.
- Global Persona Simulator — Pressure-tests product flows against diverse user archetypes in real time, automatically generating Jira tickets with clear acceptance criteria.
- UX Content Strategist — A Figma-integrated agent guiding teams on tone and terminology, ensuring consistency without slowing designers down.
03. Impact
Within the first two quarters, over 70% of active product squads adopted the AI interface as part of their daily workflow. Manual DesignOps overhead dropped by roughly 40%, particularly in handoffs, documentation, and ticket creation.
Design-to-development cycle time improved by 30–35%, largely due to tighter feedback loops and fewer translation gaps between teams.
Estimated operational efficiency gain across design and product functions by reducing reliance on external tooling and duplicated workflows.
04. Leadership & Strategy
The goal wasn’t just efficiency. It was to create a consistent, scalable way for teams to collaborate and make decisions. At the same time, the design team remained intentionally lean, supporting a growing product surface without proportional headcount increases.
Reflection
“More importantly, it shifted how the organization operates. Instead of coordinating work across tools and teams, much of the workflow became self-orchestrating.”
Teams moved faster with less overhead, and design was able to scale its influence without scaling its size. This established a foundation for AI as a first-class operational layer, embedded directly into the product lifecycle — from early concept through to delivery.