Back to Blog
Back to Blog

March 12, 2024

6 min read

Introducing six Rootly AI features: focus on the incident, leave the paperwork to us

Say hello to smarter incident management with smart summaries, mitigation message suggestions, and our new conversational assistant! 🚀✨

JJ Tang
Written by
JJ Tang
Introducing six Rootly AI features: focus on the incident, leave the paperwork to us
Table of contents

Your job is to resolve the incident, not to write pristine documentation or perform repetitive manual tasks. We’re super excited to introduce our first set of AI features, aimed at speeding up your team when an incident occurs by taking care of a lot of the writing parts for you. Automatic title generation, self-updating summaries, suggested mitigation messages, and a GenAI editor for everything you write. We’re also shipping the first iteration of our conversational assistant: Ask Rootly AI for Slack and web!

AI with Rootly is not scary

Before we get to the protein-packed part of the story, let me introduce you to how Rootly is developing safe AI. We’ve designed and implemented enterprise-grade privacy into our AI, allowing you to decide exactly how you want to interact with AI through a simple UI.

Opt-in, opt-out, or something in between

With Rootly AI, it’s all about flexibility. You have the power to opt in or out of AI features at any time with a simple toggle—no questions asked. Tailor your experience by selecting which data Rootly AI can access, from full visibility of your Slack incident channels to just the pinned messages. Additionally, you can customize which AI features are activated for your team, ensuring a balance between AI assistance and human oversight.

Safety guard rails

Rootly ensures there’s no storage or training happening with your data, not even across Rootly customers. Your data will only serve you. Plus, Rootly scrubs all personally identifiable information (PII) and sensitive data from the messages before they’re even sent out to OpenAI for processing.

Bring your own OpenAI

We provide enterprise-tier OpenAI to our customers, but for an extra level of peace of mind and alignment with your overall AI strategy, Rootly allows you to connect your own OpenAI instance. Thus, your data will be processed within your own ecosystem, ensuring a sealed data experience.

Rootly AI: More resolving, less writing

We’re implementing AI features to give you a head start when an incident pops up.

Generated incident titles

As an incident breaks, there’s no time to think about fancy titles. First of all, because you don’t know yet what’s going on. That’s why we introduced automatic incident titles through custom word banks. We’re taking this idea to the next level with AI. You can have Rootly AI suggest titles based on what’s known about the incident.

Screenshot of Rootly suggesting an incident title

You can get a suggested AI title whenever you need it; it’ll be refreshed based on the most recent understanding of the incident.

Self-updating incident summaries

Keeping an updated incident summary eases up the experience for everyone. Responders can understand what’s going on at a glance, and stakeholders can plan without having to ask around for updates. But, who’s going to keep writing summaries every time there’s an update in the incident? Well, Rootly AI, of course.

Screenshot of Rootly’s summaries

You can set Rootly AI to update the incident summary taking all the discoveries and recent highlights into account every time you need it. You can go as far as having Rootly AI automatically rewrite the summary every 30 mins via workflows.

/rootly catchup

You were just added to an ongoing incident and there’s a frantic Slack channel with hypotheses, service metrics, and references. You want to start solving the issue at hand ASAP: skip the chit-chat and get key insights of the incident through Rootly AI catchup command.

Screenshot of Rootly catchup output

Rootly AI catchup always takes the key findings and boils them down to an easily digestible format so you can get the context of the incident in no time.

/rootly resolve

Rootly AI can use generative AI to write a suggested mitigation and resolution summary for you. It will pick up the most relevant events in the incident’s history and track them until the resolution by the team.

Screenshot of a suggested mitigation

Rootly AI Editor

Grammar should pretty much be on the bottom of the priority list when everything is on fire. Yet, once everything’s over, a typo can make you feel embarrassed or confuse your colleagues in the future. Rootly AI Editor is enabled in every single text field within Rootly, so you can tidy up everything with a single click.

Screenshot of Rootly AI Editor

Furthermore, with Rootly AI Editor, you can adjust the length of your text to make it shorter or longer.

Ask Rootly AI

You can always use an extra hand when solving an incident. With Rootly AI, you can ask any details you need about the current incident and general SRE knowledge. This conversational assistant is available through Slack and web.

Graphic depicting Ask Rootly AI on web and Slack

You can ask Rootly about what’s going on with the incident and what’s been done. For example, if somebody has tried restarting the server already.

Much more AI on the way

We’re already testing more powerful features for Rootly AI. You’ll soon have access to a much more powerful AI assistant, with capabilities to access your integration’s data so you have a single interface to find out everything about an incident.

Want to see it working together?

We’ll be happy to show you around the latest Rootly AI features, feel free to book a demo!