The business problem
Two internal workflows were identified as opportunities for AI assistance: auditing support calls to help team leads assess agent quality, and generating personalized customer engagement messages intended to encourage customers to return to the product.
My role and approach
The team used AI orchestration tools to develop internal agents for support-call quality assurance and personalized re-engagement. The work was designed to expand the reach of existing teams rather than treat AI deployment itself as the outcome.
- Mapped the repetitive work and the intended users.
- Developed an AI workflow for support-call auditing to help extend QA coverage.
- Developed a separate personalized engagement workflow to support reactivation.
- Separated the KPI trees: operational hours and review coverage for QA; controlled reactivation measurement for engagement.
Outcome and value
The intended value is greater QA coverage and more repeatable personalized engagement. Quantified realized impact was not provided in the source material.
How success should be measured
Suggested planning targets such as hours saved or reactivation lift are hypotheses, not achieved results. Re-engagement impact should be tested against a credible control group; generated messages should not receive credit for behavior that would have occurred anyway.