Why chatbots are going to take over IT
A revolution is quietly taking place in human-computer interaction. Chatbots are growing up, becoming productive members of chat forums, and changing the way enterprises do business. The social life of the millennial generation revolves around text messaging, and they want the same experience in their business environment. Chatbots today can offer this experience. They interact with humans in a friendly way, without making frustrating demands on the human’s time.
What if chatbots could make employee-facing enterprise IT services much more usable? What can chatbots do for you right now, and what could they do in a few years?
Chat is becoming the norm
When was the last time you visited your local bank branch? For most common transactions, such as transferring funds, checking your account balance, ordering checkbooks, and other everyday tasks, you probably use the bank’s slick and responsive web interface, or perhaps the bank’s mobile app on your smartphone. Same with retail shopping. You don’t have to visit a pizzeria yourself; place your order online, and hot pizza will be delivered to your home within 30 minutes. This saves a huge amount of time. We don’t need to bother with the details and logistics, because the pizzeria takes care of that for us.
For consumers, the experience is really efficient. But employees in an enterprise have a very different experience when trying to conduct similar transactions at work. Need to order a computer for work? You have to log on to an archaic internal website with a quirky user interface. It asks strange questions, and often you know that it knows the answers anyway. But if you want that computer, you must supply the answers, because that’s what the workflow requires. If you need to order business cards, you might have to hop on to a totally different, yet equally quirky, site and provide the same answers you gave when you ordered the laptop. The enterprise is highly regulated and codified, but it doesn’t let you consume IT services in the same seamless and consistent way as commercial services.
Most people don't like that. Millennials especially don’t like that. They spend a lot of time on social networks, and much of that time is spent interacting with one another via text messages—WhatsApp, Facebook Messenger, Twitter, etc. At work, they enjoy similar experiences with chat rooms such as Slack, where teams form around specific topics and communicate primarily through text messages.
Team members conduct discussions, share information, and answer questions within the chat room. They get an almost immediate response, and the previous conversations in the room are always available. The transition from a social setting to a business setting is seamless. When these team members need something from their IT organization, though—or human resources, ERP, procurement, or other back-office systems—they must leave the efficient and friendly chat room and navigate to an IT application with an old and clunky user interface, to interact with a strict, rule-based, complex workflow that makes little sense to them and, frankly, doesn’t interest them—and why should it?
The good news is that this is changing rapidly. Most chat room teams today count chatbots among their members that can perform actions on behalf of the team members when they need it. A human can converse with a chatbot, much as they would chat with any other team member, and enjoy a social and convivial experience. Chatbots are increasingly able to understand natural language, respond in real time, and deliver status reports unprompted. Chatbots that are connected to the enterprise IT system can hide the complexity of the back office and gather information that the IT process requires without having to involve the human beyond what is absolutely necessary.
We’re currently using the first generation of chatbots. Even in this first generation, chatbots are beginning to mask the rules that are imposed by the back-office system and providing a consumer experience that is closer to the systems that people like to interact with daily. Humans can interrogate the system by conversing with the chatbot, which responds in a natural manner. Of course, chatbots are still bound by rules, but it’s a much more social experience.
It’s not just empowering for those who need the enterprise’s IT services; it’s also good for the IT organization itself. IT processes usually require workflow and rules engines, with a GUI on top. By using chatbots, the IT organization can leverage the chat paradigm within a text-centric framework, such as Slack, to integrate the back-end workflows to provide new services.
The users of these services are already members of the chat room, so enabling IT services through the same chat room makes them easily accessible and actually encourages team members to use the new services. There’s a short learning curve, the chatbot is always available, it’s responsive, and it will hide any complex workflow and integration that makes up the service. There’s no need for a custom UI, because the chatbot’s UI is textual. All the chatbot needs is access to the underlying systems’ APIs, and perhaps some authentication. But we know who the members of the team are and who specifically requested the service, so authentication becomes easier too.
It’s a way for IT operations to create a new engagement experience that abstracts away all of the technical noise and complexity into service offerings that deliver value to consumers, while at the same time optimizing cost and flexibility. This is also very appealing to service consumers, and is in fact becoming their new baseline expectation from service providers. Let’s not forget that developers are also consumers, and the concept of ‘ops for dev’, by which the ops team creates services, experiences and interfaces for developers, is very attractive to the developer community and streamlines their path to production.
We’re already seeing this type of chatbot today, and it’s appearing in operations chat rooms and situation chat rooms.
Engineers are currently working on the evolution into second-generation chatbots, which combine machine learning, data mining, and other advanced capabilities. These chatbots are able to analyze and extract information from the chat rooms they belong to, perform sentiment analysis on the contents of the stream, and understand what details are relevant to a specific issue. These chatbots have the potential to apply rules and workflows dynamically—by looking at what’s happening in the chat room, accessing the system of record in the back end, and understanding more about a specific user’s activity, the chatbot can enforce business process rules and workflows in a smarter and more pre-emptive way.
The second-generation chatbots are much smarter than those of the first generation. They can identify and implement business logic within the context of the operator’s activities and use that information to make decisions and complete tasks autonomously.
As an example, consider a service-level agreement (SLA) monitor that detects a breach—perhaps a transaction is taking too long to complete. A chatbot could proactively find out what code was involved in the transaction and contact other systems to get more information about that code. The chatbot is also able to understand the severity of the breach by listening to the conversation in the chat room. It will present the findings to the members of the chat room and, in the best case, explain what caused the breach, and isolate the code responsible.
This is an extremely powerful proposition, and it doesn’t involve a rewrite of the existing back-end interfaces. The chatbots communicate via a loose coupling with the IT organization’s back-end services and other supporting services, such as artificial intelligence engines, to perform analysis. Together, these will help IT organizations accelerate the delivery of services to users while minimizing user intervention and maximizing transparency.
The future of chatbots
The concept can be extended even further. Potentially, chatbots will be able to discover tools, services, and systems of record and derive the rules and behaviors that connect and govern them based on the activity that it sees in the chat room. After all, this interaction between people, and between people and systems, is where knowledge and experience are created.
An example might be a chat room that discusses and resolves defects. Chatbots can listen in to the way the defects are handled in the chat room and learn what they need to do in order to log a defect, model or describe its behavior, find out what code it involves, and even figure out how to resolve it.
One of the primary effects of continuous integration (CI) and continuous delivery (CD) is that automation can be leveraged at a greater scale. By introducing chatbots into the CI/CD chat rooms, you can further increase your development velocity and reduce your delivery time.
IT organizations will enjoy similar benefits. By looking at the IT business as a value stream, you can articulate the value of the processes in the organization. Once you can measure the value, you can demonstrate that you can deliver faster and reduce your customers’ wait time.
Chatbots are going to take over the IT world
Even in the first generation, chatbots have tremendous benefits to IT organizations:
- The experience of interacting with IT is changing to become more sociable. The chatbot gives you the tools that you need, when you need them, and transforms the experience of interacting with the IT organization to one that customers and users will actually enjoy, due to its social nature.
- You can deliver value to your customers much more quickly, because the chatbot focuses on the integration of systems and use cases rather than on developing custom user interfaces on top of workflows and business processes.
- It clarifies the vision for the enterprise beyond the IT organization as you start to implement this approach across the value chain. The introduction of chatbots to HR, supply chain, CRM, finance, and other parts of the organization will drive efficiency, and interaction with these systems becomes more natural.
Chatbots are a better way to do workflow. They create systems of engagement that are text driven, rather than UI driven. Unlike traditional interfaces, the user isn’t forced to learn the system. It’s the other way around—the chatbot learns the user and suggests and directs the user to services and capabilities that match their behavior. This integration of people, processes, services and systems in an efficient and natural way has a natural appeal to millennials. As chatbots evolve, they will continue to disrupt IT, and other parts of the business, for the better.
Image credit: Flickr