What are some use cases you are seeing for Watson Orchestrate and automation?
If I look at the labor side of things, it’s not going to be a generic one—meaning like you create a process mining tool. A process mining can be used across any processes to mine the logs to say, ‘OK, this is the process that’s most vulnerable to automation, and let’s automate it.’
You create a workflow tool, you can use a workflow tool against any discipline. I think a digital employee is not that way.
The reason I say this—there are specific skills, meaning a digital employee for HR is going to have different skills than a digital employee for finance, will have different skills than for a digital employee for insurance, for example.
So it’s going to be a vertical-based skill set that you have to create. And these skills have been with people for a very long time, so it’s about extracting the skills and also bringing AI into it. … I would say digital labor or automation based on AI is where the future is going to go.
Now to answer your question—so let’s think about it. Today, an NLP [natural language processing]-based front end exists. I can have a chatbot when a request comes in. But it could come through an email, it could come through a voice.
It’s all about recognizing that and saying, ‘OK, what is the end user asking?’ So I could be saying, ‘Get me the latest report or my sales forecast for the next 30 days.’
Now, all of a sudden, that is a request that comes in, and then it’s taken and looked at from a sales perspective to say how many skills do I have to answer that query? That’s the first thing that needs to happen. And that’s why I said, even these skills have to be put together in a way that’s very vertical or industry-specific.
So from a sales perspective, maybe 10 skills could raise hands and say, ‘Hey, I have that skill to answer that query.’
And then it’s all about bringing the confidence level to say, ‘OK, you said—what is your confidence level? I have a 96 percent confidence level that I can answer that query.’
But it’s not just about answering that query because in the back end you’ve got to connect to a sales force system, extract that. You need to have email skills because you might have to email it somewhere.
And then it’s not only an email skill. That’s where another role like iPaaS (Integration Platform as a Service) comes in because now you have to connect to different systems and be able to extract data from there or insert data into there and then generate it as a PDF, send it as an email, and make sure it all goes. And that’s why the sequencing matters.
So to answer your question, I would say I can take any example like Watson Orchestrate is being used in IBM HR today to decide the bonus level based on the performance of an employee.
I mean, think about it, we had to have someone go do that manually … but end to end, it’s taken care of. And we are trying to expand it across IBM HR in many disciplines.
So I would say it depends on the discipline, but we are already using it. My goal is that over the next three to five years, democratize it. Meaning make it available to everyone you know.
I’m a mom-and-pop shop, I want to have a digital assistant to do all the mundane tasks that I do but I don’t have the resources to go hire someone. How wonderful it would be for me to have as a printing shop or a bagel shop to be able to use what an enterprise has the capability and the resources to go do it.
So my goal is not only to bring Watson Orchestrate for the enterprises, but how do we make it easy and simple enough that you can easily build a skill, you can put it into a catalog and then Watson Orchestrate can see all those skills and execute and sequence it?
So if you ask me, one, bring that to the enterprises at the level of scalability, security and all those things on a vertical aspect.
But two, democratize it, meaning bring it as a service that anyone, everyone can come and use it and build their own digital employee.