Feature Highlight

Cortex Notifications: Stay up to date while staying in flow

Cortex’s notifications were thoughtfully designed to minimize distractions for engineering teams. In this post, we’ll outline what notifications engineers can expect and how they can stay in flow for longer with Cortex.

Ryan Garrett
January 23, 2024

The problem with notifications

Notifications are designed to be annoying. Think of your phone buzzing in a quiet room: it demands your attention, lighting up your screen and making noise so you look at it. A notification is supposed to pull you away from whatever you’re working on. They can be useful, but they can also be a nuisance.

For software engineers, the disruption caused by notifications is more than a nuisance. A notification pulls them out of flow, taking attention away from the task at hand. And, once they return to what they were working on, they need to load the codebase back into their ‘mental RAM’ to continue, which takes time and significant mental effort. That’s why it’s so important for engineering teams to carve out focus time, and avoid distractions from other teammates.

With modern engineering teams using dozens of different software and tools, notifications done wrong can quickly decrease developer productivity. On the other hand, developer tools that provide targeted, prioritized, and interactive notifications can help developers schedule their work and stay on track against important deadlines.

Prioritized Developer Homepage

As an engineer-founded company, we’ve spent a lot of time thinking about the best way to deliver information to engineers.

While other catalogs and IDPs provide devs with a wall of information about software upon log-in, Cortex curates the information shown without limiting information that can be accessed. When an engineer logs into their personalized Developer Homepage, they’ll find a list of tasks prioritized by deadline and importance, eliminating the work needed to sift through all these task notifications manually. Cortex also centralizes relevant information for engineers into one place to avoid the need to context switch between different tools. Information from Jira, Pagerduty, and Git will be surfaced in Cortex, letting engineers focus on their tasks at a high-level without flipping through dozens of tabs to get the information they need.

Notifications in Slack

All of this information surfaced in the Cortex portal is valuable, but Cortex can also meet engineers where they already are: Slack. With Cortex’s bi-directional Slack integration, engineers can not only ask for specific information about a software component, they can also receive highly-relevant notifications about changes and updates. This two-way integration sets Cortex apart from typical one-way notification systems and makes it easier to access the valuable information in Cortex. And rather than spamming engineers, Cortex will roll up these updates into weekly digests and target the engineers who need to receive them.

Critical notifications during on-call

For software engineers, one of the most stressful times is being on-call when there’s an incident. Many engineers have horror stories of a particularly gnarly incident, often exacerbated by not being able to find critical information fast enough.

During an incident, the on-call engineer needs to retrieve context to help them resolve the incident, such as:

  1. Who owns this software component? If I have questions, who should I talk with?
  2. What changed? This software component was working yesterday, so what happened here (or in an upstream dependency) that caused this problem?
  3. Where can I find more information about this software component? Does this software have a README, runbook, or other documentation that might help me debug this incident?
  4. What tooling is associated with this software component? Do I have dashboards or logs that might provide insight into the root cause?

Without Cortex, an engineer responding to an incident may need to log into a number of tools like Datadog, Pagerduty, or Confluence — and then dig around to find the necessary logs, runbooks, dashboards, discussion, and documentation. No one enjoys digging around for docs that may or may not exist, especially when there’s a critical outage at 3am.

Cortex designed the new On-Call Assistant with this in mind. When an incident happens, Cortex neatly packages up all relevant information and notifies the on-call engineer in Slack, giving them all the context they need to resolve the issue:

Whether the on-call engineer needs to see when a service was last deployed, what the logs show, or find relevant documentation, they can directly query Cortex in Slack and avoid jumping between a sprawl of tools, saving precious time when responding to a critical incident.

Deadlines without the dread

While one-time incidents are demanding because of their immediate urgency, the opposite situation also poses challenges for modern engineering organizations: long-term continuous improvement programs like security, maintainability, and quality. Whether your team wants to incrementally improve the quality of your existing software or simply ensure that all software components have a README, these long-running programs can quickly become forgotten about or deprioritized when other work pops up.

Cortex’s Scorecards let you set standards across the organization and rely on Cortex to track incremental progress towards achieving those standards. If you want to establish more urgency around a particular standard, you can create an Initiative to set a deadline, so engineers who are responsible for that work can prioritize it appropriately. Engineers will see these tasks (and their deadlines) tracked in their Developer Homepage, and will be notified in Slack as the deadline approaches to ensure the work doesn’t get dropped. This means fewer pings from an engineering manager or program manager — and a happier team overall.

TPM in a box: headache-free migrations

One of the biggest frustrations that engineering teams face is performing migrations—to the cloud, away from the cloud, onto new software, and away from those EOLing, etc. It might be easy to get to 60% across the team, but getting everyone on the same page is seldom light work. Where are the stragglers? Why haven’t they migrated? Do we need to file an exemption or have folks just been away on vacation?

These are surprisingly hard questions to answer. Most teams do so by having someone (often a TPM), manually monitor the status of projects and either incessantly ping folks directly, or worse—blast entire Slack channels, spamming engineers who aren’t responsible for (or have already completed) the migration.

There is a better way! Many Cortex customers have slashed migration time by setting up a Scorecard and accompanying Initiative (with a deadline for the complete migration). Cortex will then do all the heavy lifting of notifying the relevant engineers — and only the relevant engineers — about their pending work. LetsGetChecked shaved 8 months off their planned Kubernetes migration by using Cortex Initiatives to help developers stay focused on high-priority tasks. Other customers have also seen significant time savings by offloading follow-up notifications to Cortex during time-sensitive AWS EOL migrations.

Distraction-free notifications

Because Cortex notifications are thoughtfully designed, engineers can remain in flow for longer, but also have relevant information pushed to them (whether that’s urgent incident context delivered immediately, or FYI updates delivered weekly)

If you’re new to Cortex and would like to learn more, request a live demo or check out our interactive demo to see how Cortex can help you cut noise for developers and improve your developer experience.

Feature Highlight
Ryan Garrett
What's driving urgency for IDPs?