← Case files
Case File 01Banking

TD Bank

Building an operating system for authentication analytics and team visibility

Senior Business Information Management Analyst · 2025–Present

Structured Jira, Confluence, intake, dashboard documentation, and stakeholder reporting for a team working across authentication, fraud, analytics, and digital trust.

TD Tower, Toronto
TD Tower — Toronto
~30% reduction in stakeholder back-and-forth
10+ request types unified into one intake
6 dashboards documented and governed
Reporting reaches EVP level
Signal

A high-stakes team running authentication, fraud and analytics had no Jira, no structured intake, and requests arriving as scattered emails.

System

There was no operating system around the team — work was hard to track, hard to prioritize, and invisible to leadership.

Decision

Treat the team itself as a product. Prioritize intake, documentation, and reporting over any single dashboard.

Build

Stood up Jira to TD standards, built a Confluence intake that auto-creates structured tickets, authored the team's knowledge base, and introduced a weekly reporting rhythm.

Outcome

~30% reduction in stakeholder back-and-forth, EVP-level visibility into the team's work, and a clear surface for AI-agent and authentication-optimization initiatives to land on.

Learning

Instrument the operating system itself — time-to-resolution by request type, not just request volume — so the case for process changes is quantitative, not anecdotal.

Detailed file

Context

TD's Customer Authentication Strategy & Performance (CASP) team sits at the intersection of fraud prevention, customer experience, and authentication analytics. Data scientists and business analysts monitor authentication journeys, optimize rules and models, and are building AI agents for authentication optimization and simulation. When I joined, the team had no Jira, no structured intake, and no formal operating model.

Jira operating model from zero

Initiated the Jira setup within TD's organizational standards. Created the project, defined all fields and issue types, organized work by pods and workstreams, linked tickets to partner team boards, introduced weekly one-on-one cadence sessions with each team member.

Automated request intake

More than 10 different request types were arriving via unstructured email. Designed and built a Confluence intake form that automatically creates a structured Jira ticket on submission — entirely self-initiated. Estimated to reduce stakeholder back-and-forth by ~30%.

Confluence knowledge base

Built a comprehensive Confluence space — team overview, dashboard documentation, project docs, onboarding guides, governance and audit-readiness documentation, alert playbooks. Used by the entire organization including the Executive Vice President as the authoritative reference for the team.

Stakeholder visibility

Weekly progress reports for senior management. Self-initiated a quarterly newsletter — later expanded org-wide to improve CASP's visibility and internal brand.

Real product work isn't always a feature. Sometimes it's building the conditions where the actual work — including AI agents and analytics — can ship, get adopted, and be trusted.
What I'd do differently: Instrument the intake system earlier — tracking time-to-resolution by request type and pod, not just whether requests were submitted. That data would have made the case for process improvements quantitatively, not just directionally.