Blameless’ Incident Resolution chatbot is getting a makeover. We’re excited to share how this change came about, what the revamp includes, and how Blameless customers can get the most out of it.
From Hack Day to production
It’s Hack Day, Winter of 2021, and Software Engineer Simone Salman has an idea. She wants to reimagine our Slackbot. This sort of initiative is one of many reasons we love Hack Days over here at Blameless. Team members are given a chance to shine and think outside the box. After Simone’s demo, the team cheers, “Ship it!” and her bot beautification project becomes a production reality.
This project was important to Simone because, as she says “The idea for the bot work primarily came to us because we wanted to be able to render tasks and incidents in a sleek way. The goal was to make it less bulky so that users can spend more time thinking about resolving the incident rather than managing the incident. We hope that the new task lists allow customers to take meaningful actions during the incident in a simpler manner.”
Simpler, sleeker. These were the words our team kept in mind while building this. Our old Blameless bot (while well-loved) was a bit… noisy. During an incident, teams want their channel to be as clean and clutter-free as possible. Too much, and it can be overwhelming to find the information necessary to move forward.
Also, while our commands were straightforward, they could be a little tough to remember, especially in the thick of an incident where an engineer’s mind is on so many other things. So another important part of this upgrade was highlighting helpful commands to get the incident started. This can reduce the cognitive toil required while participating in an incident, and makes Blameless even easier to use for those just learning the tool.
Additionally, with the help of Design Lead Alice Tung, we improved the UI/UX of our bot. Alice explains why this is important: “Although the possible UI and layout was highly constrained in Slack, design could still have a big impact through content, as the users needed the right information at the right time! The solution was not to send fewer notifications to Slack (they were all important), but try to organize important information that is easy to digest.”
To round out this dream team, Product Manager Divya Venkatachari stepped in to add the finishing touches and plan for a smooth launch. Let’s see what these improvements look like in action.
The Blameless bot reboot
Here are the improvements we’d love to highlight about the new Blameless bot:
Updated Incident Summary: The incident summary has been updated to display a more concise rundown. The new summary has improved UX/UI design and shares the incident summary, severity, status, and type as well as timestamps and the people involved.
Automated shortcut message: As you can see above, we also added a message that gives users access to documents and shortcuts to help automate commands. This lowers the toil for responders.
Incident help suggestions: When launching a new incident, the Blameless bot will now provide suggestions for commands. The bot will direct the user to the list of commonly used "slash" commands and provide a link to the web site where the commands are all listed.
Task list enhancements: In addition to creating an expanded checklist, we’ve also allowed tasks to be checkmarked within the Slack UI for a smoother user experience. To limit the noise within the Slack channel, now only task owners can see their assigned tasks. When creating a new task, it will appear to the owner in readable format and updates the person’s main task list.
Lastly, rather than change task status from a drop down, tasks are crossed off as they are complete. This makes the previous Blameless commands “_mark task as pending” and “/blameless complete task” irrelevant. Instead, users should use the command “/blameless show tasks” and cross off completed items from the checklist that appears after the command.
These improvements are just one of the many ways Blameless works to make engineers’ jobs easier. By improving our UX/UI, streamlining commands, and reducing noise, we’ve eliminated toil from your incident response process.