Key Results

  • 3x faster time to resolution
  • Improved workflow & triaging
  • Custom Alerts to reduce noise

SDK

Javascript, Python, Go, React, Android, Cocoa

Solutions

Error Monitoring, Mobile Application Monitoring, Performance Monitoring, Context, Custom Alerts

Related Content

Logging in Python: A Developer's Guide

Sentry + PagerDuty Integration

Understanding the Performance Impact of Generated JavaScript

By filling out this form, you agree to our privacy policy. This form is protected by reCAPTCHA and Google's Privacy Policy and Terms of Service apply.

Share

Share on Twitter.
Share on Facebook.
Share on LinkedIn.

How Virta Health resolves critical issues 3x faster by monitoring their frontend, backend, and mobile services with Sentry

About Virta Health

Virta Health offers a new care model for type 2 diabetes, helping patients reverse their disease while reducing or eliminating the need for medications like insulin. To do so, Virta created a virtual medical practice where providers guide patients through nutrition therapy and lifestyle changes to normalize blood sugar. Over time, Virta clinicians can then safely de-prescribe medications. In this model, patients actually get healthier, flipping the idea of diabetes as a chronic, progressive disease on its head.

At one-year of Virta’s peer-reviewed study, 60% of participants achieved diabetes reversal and 94% reduced or eliminated insulin usage. As further evidence of its success, Virta scaled funding and headcount significantly over the past year, and experienced a 3x growth in customers.

Transforming manual error tracking into a scalable process

Virta has experienced dramatic growth in the last few years—more than tripling its customers. Stephen Akinyooye, Senior Software Engineer, recalls that as their team and product grew, Virta wanted a scalable solution to help them proactively respond to issues affecting their growing customer-base.

Before Sentry, the Virta engineering team would discover issues through customer support tickets. And while they had a good handle on identifying exceptions, they would manually triage and handle errors using log aggregation services. This process required each engineer to take turns reviewing logs and triaging—which was inefficient and reactive.

As we grew as an org, the amount of services and applications that we managed grew as well. The need for a centralized place and process to monitor and triage the health of our apps and services was becoming more critical. We wanted to respond to issues before they were affecting our patients.

Stephen Akinyooye, Senior Software Engineer

Requirements

Concerned with reliability and stability, Stephen and his team wanted a solution that would prioritize, contextualize, and alert them of errors in real-time to support their growing team and product.

Given their previous challenges, integration with their existing tools such as PagerDuty and Jira was essential to streamline and complement their workflow. They were also looking for robust platform support to monitor their projects (built on React/React Native, Python, Go, and others).

In their search for a solution, Virta chose Sentry. “Sentry was the market leader that fit our requirements,” said Stephen.

Building a streamlined code observability practice

Since working with Sentry, Virta has implemented Sentry across their entire stack from web, to their React Native app, and their backend.

“Sentry has become a go-to tool for me and my team. It enables us to triage errors in a more structured and efficient way. We can now contextualize issues and access comprehensive details to figure out what’s going on faster,” explained Stephen.

As Virta was implementing Sentry, they took the time to understand how they could leverage Sentry’s full suite of capabilities within their existing workflow and tooling.

“With Sentry, you can build a process such that you’re not just responding, but triaging, managing, and retrospecting around your systems and production environment.”

By incorporating Sentry products such as Release Health and Discover into their workflow and integrating it with the tools they already used, Stephen and his team are able to expedite resolution time by 3x with the context that Sentry provides, and the seamless collaboration that follows.

For instance, using Metric Alerts, Sentry’s rule-based alerting, with our PagerDuty and Jira integration means staying on top of issues as soon as they occur—and increasing collaboration across teams.

Virta pushes updates to its software frequently. They use Release Health to track how every release is performing. In the event there’s a critical issue, they leverage suspect commits to learn which commit caused an unhealthy release. “Release Health alongside Suspect Commits makes it easier for us to triage issues by providing essential context to our team, showing the full stack trace and codepath that caused the error, who released it, and clues to solve it,” said Stephen.

And finally, Discover, our query engine—which the team found especially useful as they changed underlying infrastructure—allows for a coarse-grain insight across issues and projects.

“With Sentry, I can find out details around device type or version, as well as access the stack trace to identify the exact error and replicate and diagnose what happened faster. All of those attributes are automatically recorded and the most important attributes can be queried. And you can set custom tags and base queries off of those as well,” explained Garrett McCullough, Senior Mobile Engineer at Virta.

The Virta engineering team hoped for efficiency, familiarity with their environment, and a tool that was easy-to-use and maintain.

“From an engineer standpoint - [we want] tools that are easy-to-use and maintain. We’ve been able to build a process around Sentry and its usage that supports us, and it’s become a key piece of our stack,” said Stephen.

What’s next?

As the team and product grow, Virta’s engineering team is beginning to set more application and service performance objectives and looking to consolidate their ecosystem. Gaining insight into deterioration of performance and efficiently reconciling those performance issues with errors is essential.

“While we have a spectrum of tooling that allows us to inspect some of these performance characteristics, the value Sentry delivers in tying together an ecosystem into one tool makes it easier for our engineering team to see performance issues and solve them faster,” explained Stephen.

Results

In setting performance objectives and tracking progress of their growing team, one of the most important metrics is incident resolution speed. Amazingly, with Sentry, Stephen and his team were able to identify and triage issues 3x faster than before.

Sentry is more than a canary in the coal mine. When we see errors come up, when I’m seeing something that I’m nervous about - I keep my eye on Sentry.

Stephen Akinyooye, Senior Software Engineer

A better experience for your users. An easier life for your developers.

© 2024 • Sentry is a registered Trademark
of Functional Software, Inc.