# How ezCater uses Level AI

**13%**: “13% decrease in its overall call handling time”

As published on [thelevel.ai](https://thelevel.ai/case-studies/ezcater). Captured by usedby on 2026-10-08.

- Company: [ezCater](https://www.usedby.ai/companies/ezcater.md)
- Tool: [Level AI](https://www.usedby.ai/tools/level-ai.md)
- Teams: Contact center agents, Supervisors, Quality Assurance

## What the story says

ezCater's contact center uses Level AI's Agent Assist to surface relevant knowledge base articles to agents during live calls. It also uses Manager Assist to give supervisors real-time visibility into calls, and Auto-QA to score quality on customer experience and solution-oriented metrics.

Summary written by usedby from the source page, in English. The figures are those of Level AI and ezCater, not ours.

> Agent Assist, with its real-time knowledge access, helped ezCater improve the day-to-day experience of its agents and customers. Since implementing Agent Assist, ezCater has seen a 13% decrease in its overall call handling time.

- **23%**: “23% reduction in call hold time during peak lunch hours”
- **94%**: “94% of calls served within 30 seconds”

> Level AI not only allows us to assess where we currently stand in terms of call quality, but also provides insights on how we can continuously improve. Reviewing our calls helps us identify strengths and areas for growth.
>
> Arianne Villarosa, Customer Service Specialist

> Be really thoughtful about the training for both agents and leadership. Explain how they can leverage these tools to enhance their experience at work for successful adoption
>
> Michaela Conserva, Senior Manager of Quality Assurance

## What usedby checked

We compared the story with its live page on 2026-10-08.

- Checked: the figure 13% is printed word for word on the page, near the name of ezCater.
- Checked: the passage quoted above is copied word for word from the page, near the name of ezCater.
- Checked: ezCater uses Level AI. Confirmed line. Latest check across sources: 2026-10-08.
- Not checked: the result itself. We quote it; we did not measure it.

---
Source: https://www.usedby.ai/case-studies/ezcater-level-ai · How we check: https://www.usedby.ai/methodology