Scaling event support without scaling the team.

Situation

FIFA World Cup-related operations created a concentrated increase in driver and customer support demand. The problem was not simply adding a chatbot: the workflow needed useful knowledge, clear operational boundaries and a reliable route for cases that required a person.

My responsibility

I owned the operational implementation: shaping the support-agent workflow, preparing its dedicated knowledge structure and getting it into use. My focus was on making the system useful to the support operation rather than treating it as an isolated experiment.

What changed

The agent was deployed to handle the defined support scope, with knowledge organised for the production workflow and escalation left available where automation was not appropriate.

I worked with the people closest to the operation to make the workflow practical, and used their feedback to identify where the agent needed clearer knowledge or a more cautious hand-off.

Results and measurement context

The workflow handled thousands of queries and recorded 76% observed CSAT. During the demand spike, additional human-agent scaling could be frozen.

What this demonstrates

I can take a support problem from operational framing through knowledge preparation and deployment, while keeping the outcome tied to service quality and the point at which a human should remain involved.