IT Operations Management (ITOM)

ChatOps – ChatBots and Enterprise IT

ChatOps – ChatBots and Enterprise IT


2016 has been the year of the ChatBots (bots) disruption. Adopting a notion that has been common in China for the last few years, we’ve seen software vendors such as Google, Facebook and Microsoft announcing their new releases of their chat tools, which are powered by bots. We’ve heard of CEOs of big companies betting on ChatBots. We’ve seen Venture Capitalists betting on startups that create bots, and on the startups that create technologies to create bots.

But why are bots so interesting?

ChatBots are seen as a wave of disruption—on top of the disruption that mobility introduced to our lives. They are making services more accessible than mobile apps do, and they allow users to consume a service by simply starting a chat with a bot within their chat app—which they use throughout the day anyway.

Most of the new hype is focused on Consumers or Small Businesses, though.

In this post I will explain how ChatBots can transform Enterprise IT.

We’ll begin this journey by understanding the Consumer space, continue by understanding chatbots for teams, take a turn towards the enterprises and finally arrive at Enterprise IT.

Buckle up and enjoy the ride!


ChatBots helping Consumers

The notion of bots in the consumer space is dominated by two main concepts: Artificial Intelligence and Accessibility.

Artificial Intelligence is a cornerstone in conjuring smarter bots that can better “understand” our needs, can find us things that we desire, or make a recommendation based on learnings from our buying or usage patterns. This is the realm of personal assistants, such as Siri, Cortana, Google Assistant and others. There are also many new companies offer their services (shopping, entertainment, travel, and more).

Advancements in Artificial Intelligence also help satisfy the high expectations we have from consumer-grade engagement experiences. Most of us are only willing to interact with bots, if the experience is rewarding. A variety of vendors are offering Natural Language Processing services that allow for creating bots that can converse in plain English (as well as many additional languages). This is the experience users expect when chatting within a chat app. In the next few years, we can expect to find an increasing number of chat bots that are gradually more pleasant to interact with. However, NLP, although significant is not mandatory for creating a robust user experience. The Chinese market has found alternative ways of producing bots that get the job done.

While Artificial Intelligence can be used to power many forms and mediums of technology, the novel thing about bots is that they significantly increase the accessibility and the immediacy of the services. This is a wave of disruption on top of the first disruption that mobility has introduced into our lives. Mobile apps have made services more accessible and ubiquitous in our lives, yet consuming a new service comes with complexity—such as downloading a new mobile app from their App Store and further cluttering their phone. With Chat Bots this friction is eliminated. Consumers can now simply interact through the service bot, and shop, book a taxi, order a pizza, or find a movie to watch later on tonight.

So, in the consumer space, where one-to-one interactions of a person with a bot is the dominant set up, the Chat Bot innovation is mostly centered on Artificial Intelligence and Accessibility.

Let’s now understand how Chat Bots can provide value to teams.


ChatBots in teams – Powering up collaborations with ChatOps

ChatOps 1.png


The use of bots within teams and organizations is significantly different than in the consumer space. Here we are looking at one-to-many interactions between a bot and the team, or even Many-to-Many type of interactions where several bots are used by an entire organization.

Team collaboration is a space filed with variety of useful technologies to improve effectiveness and efficiency. Varying from mature technologies such as audio conferencing, through voice-over-ip, video chats, mobile devices, messaging tools, and recent arrivals such as the modern wave of collaboration tools (Slack, HipChat, FlowDock, MatterMost and Rocket.Chat are a few to mention).

But while all of these technologies are used to improve the medium of collaboration between people, they keep the collaboration siloed from the systems in which people actually do their work. Chat Bots help bridge this gap, and they add the context to the collaboration. The emerging practice here was named ChatOps by GitHub, and has been gaining momentum in the team collaboration space since around 2013, when GitHub shared with the world how they powered their own Dev and Ops processes with bots.

Consider the following example depicting how R&D teams typically cope with identifying the cause of a failing build, prior to the introduction of ChatOps. It many times goes like this:

  • A build fails
  • Several developers were involved in the latest code check-in. Additional QA engineers are responsible for the failing tests
  • The build manager starts chasing people involved, via emails or phone, to figure out who’s responsible for the failing build
  • The developers in turn may need to have a discussion with QA, to understand why certain automated tests are failing
  • Each person that joins the discussion needs to be brought up to speed on everything that happened so far
  • Most of the people in the R&D team do not have visibility to whether anyone is addressing the build issue
  • After the build is finally restored, not much data has been captured to help in future incidents when they arise

The ChatOps practice is all about introducing the systems into the collaboration, via bots*.  Alerts, tickets and additional information from our Development and Production systems are setting the context for the conversation. This is accomplished by opening a dedicated chat room, inviting the right people, and putting vital information in front of their eyes.  Further, data from the systems is powering the conversation and helping to drive more informed decisions that cross people, teams and skills. Actions that are performed by conversing with the bots are all logged within the conversation timeline, allowing all participants and stakeholders to gain visibility to everything that has already been tried. When the build is restored – the entire dataset from the exercise is captured and logged for any future audit or retrospection.

It might look like this:

Daniels ChatOps.png

* Image courtesy of HPE ALM Octane R&D  and their bot

When used for build management within R&D or DevOps teams, organizations see significant improvements in the time it takes to restore failing builds, as well as improved collaborations across teams that commonly operate in siloes. In general, organizations see happier and less stressed build managers, project managers, R&D and QA engineers when practicing ChatOps.

So the use of ChatBots for Teams is quite different than the use of Chat Bots within the consumer space. It is about adding value to groups of people that are joined together around a certain task. It is about a landscape that expects multiple bots to co-exist and complement each other. When comparing to the consumer space, accessibility remains a prime priority and factor, while Artificial Intelligence although useful and important becomes lower in its priority.

As an example, the majority of ChatOps implementations practiced today do not require their bots to support NLP, as sometimes developers and IT people even prefer their bots to be deterministic on the commands that they accept.

Now that we have explored how Chat Bots can help Consumers as well as small teams, in the next blog post we will look at what Chat Bots for Enterprises.

*an alternative method is by directly integrating the IT systems to the collaboration tools. I’m discussing the pros and cons of this method in the third part of this post series.





  • operational intelligence
About the Author


Oded leads the ChatBot and ChatOps Strategy for the HPE Software IT Operations Management business.. His background is in the DevOps and IT Service Management domains.