FlipCX Cut Observability Costs by 50%+ with groundcover

About Flip

Flip provides purpose-built Voice AI solutions for industries where phone interactions remain essential.

Their platform automates tasks such as:

• Order tracking and subscription management for e-commerce
• Non-emergency medical transportation coordination
• Appointment scheduling and billing automation in healthcare
• Taxi bookings and trip management

Unlike generic AI assistants, Flip builds industry-specific AI systems, supported by dedicated teams.

This allows them to deliver higher-quality automation. Flip’s infrastructure includes:

• Long-running WebSocket phone calls
• Speech-to-text pipelines
• Multiple AI providers
• Kubernetes workloads
• Dozens of external APIs

Many of these systems run 24/7, which makes observability essential.

“Our platform runs 24/7, so observability is critical. We try to find issues before customers even notice them.” 

- Joe Garlick, Lead Architect, FlipCX

The challenge

Observability Costs Were Getting Out of Control

Before adopting groundcover, Flip relied primarily on Datadog for logs, metrics, dashboards, and monitoring. As the company scaled, observability costs increased dramatically.

“Our Datadog bill had climbed. It’s hard to justify that for logging when it doesn’t directly affect the phone call experience.” 

- Joe Garlick, Lead Architect, FlipCX

Retention limits compounded the problem. Flip could only keep logs for 7–14 days, which meant many incidents lacked historical data for investigation.

“Customers often report issues weeks later. Before groundcover, we just didn’t have the logs anymore.” 

- Joe Garlick, Lead Architect, FlipCX

At the same time, Flip’s architecture was becoming more complex. AI services and multiple APIs were involved in every phone call. If one of those services failed, the call could fail. 

“When you start telling engineers to log less because it's too expensive, something's gone wrong.” 

- Joe Garlick, Lead Architect, FlipCX

A Rigorous Evaluation: Testing 20–30 Vendors

Flip didn’t rush the decision to replace Datadog. Joe Garlick launched an extensive evaluation process and tested 20–30 observability platforms. Instead of relying on vendor demos, the team ran real proof-of-concept deployments. Production workloads were routed through each platform.

“We didn’t rely on demos. We pushed real traffic through the systems to see how they behaved.”

Flip’s platform generates enormous log volumes. At the time of testing, it was producing over 10 TB per month. Many platforms could not handle that scale. Some tools became extremely slow when searching logs. Others crashed entirely.

“I started pumping logs into one vendor’s system and eventually crashed the entire instance.”

Another non-negotiable requirement was fast free-text log search. Customer support tickets often contain only partial clues about a failed call.

“Sometimes a ticket just says the caller mentioned ‘McDonald’s’ within a two-hour window. Being able to search that instantly is a hard requirement for us.”

Many observability platforms struggled with this workflow. Others had usability problems.

“Some tools had a fast backend but a terrible UI. Others had a great UI but were slow. groundcover was the first platform that got both right.”

Immediate Visibility with eBPF Observability

Once Flip deployed groundcover, the team immediately gained deeper visibility into their systems. groundcover’s eBPF-powered observability automatically captures infrastructure and application activity without requiring manual instrumentation.

Engineers can now observe:

• HTTP requests
• MongoDB queries
• MySQL activity
• API errors and failures
• rate limiting events

Within the first day, Flip engineers discovered issues they had never seen before.

“Within the first day we were already spotting issues we didn’t even know existed. We immediately started seeing rate limits, failed requests, and errors that weren’t logged properly before. That was pretty game-changing.”

From Logs to Full Observability

Before groundcover, Flip relied heavily on logs to understand system behavior. groundcover introduced true application performance monitoring. Logs, traces, infrastructure metrics, and Kubernetes visibility are now connected in one place.

Engineers can:

• Trace requests across services
• Inspect container performance
• Analyze API dependencies
• Monitor system resource usage

This visibility is particularly important for Flip’s AI-powered workflows. groundcover allows the team to monitor these dependencies and respond quickly.

Observability Across the Entire Company

One of the most surprising outcomes of adopting groundcover was how widely it spread across the organization. Today, 30–35 employees actively use groundcover each month, out of a company of about 50 people. That means nearly 70% of Flip’s workforce interacts with observability data.

“Nearly everyone in the company uses groundcover now, not just engineers.”

Teams using the platform include:

• Engineering
• Customer success
• Data teams
• Sales
• Implementation specialists
• Leadership

Customer success teams frequently investigate incidents before escalating them.

“Our customer success team can look up a call, see the error, and sometimes solve the issue without even involving engineering.”

This significantly reduces investigation time and allows engineers to focus on building new capabilities.

Longer Retention, Lower Costs

groundcover also solved Flip’s retention and cost challenges. Before switching: Logs were stored for 7–14 days.

With groundcover: Logs are stored for 30 days by default, with the option to extend retention further.

“Moving from two weeks of retention to 30 days was a huge improvement for us.”

At the same time, Flip reduced observability costs dramatically. Monthly spending achieved more than 50% savings, while gaining deeper observability capabilities.

Why Flip Chose groundcover

After evaluating dozens of observability platforms, Flip chose groundcover because it delivered the right combination of performance, usability, and operational flexibility.

“groundcover is the full package. The UI is super clean, the backend is snappy, and keeping everything on-prem means less compliance headaches. On top of that, their team is always there to help and they're constantly shipping features that actually move the needle.”

For Flip, groundcover provides:

• Faster debugging
• deeper infrastructure visibility
• company-wide access to observability
• longer data retention
• significantly lower costs

Most importantly, it helps ensure that every automated conversation runs reliably, no matter how complex the system becomes.

Want full visibility into your most critical systems without the cost and complexity?

Try groundcover free today.