How to get started with AI & the bot movement
With all of the advancements made thus far, we are still very early in the days of AI & ML. With Tesla and Boston Scientific it is easy to see these wild achievements and wonder what is next on the grand stage but equally interesting to me are what I am considering more micro wins in the day to day business landscape.
Even while I type this post, AI is watching me. WordPress has a little tool in the top left corner of my screen telling me that my heading sucks and I need to do a better job titling this blog if I want someone to read it. Such a simple application with no direct revenue benefit but an interesting tool that is a small step to round out their entire platform. (39/100 Red = bad title)
On a day to day basis, myself and my team are pulled into organizations to help make an impact with their technology resources. Make more out of less and oh by the way, make sure to squeeze in some innovation into the mix where possible. CIO’s & IT leaders are juggling more than ever to keep up with the business and while they feel like they are drowning with Cloud, let’s just throw in an AI topping into the mix to really shove them down the deep end. So other than a pity party, I’ve been thinking about how I can assist these folks to get some wins, show the business we IT collectively are moving in the right direction while also not sinking the ship and I think I have a few ideas.
First- how I frame AI & ML in my head. When I think of AI & ML, I am thinking about integration platforms working with RPA, bots & software to basically operate itself. I am no deep learning expert nor do I want to be right now, we can hire for that, however, I am always curious. This entire domain is more than just some data analysts whipping up some cool data correlations and then serving it up to the business on a recurring basis. (I’ve been told that is AI)
How do we get started? – This is the big question and I think I have some idea of where to start. My line in the sand is going to be getting the quick win and driving belief and adoption in this domain. How do we do that? We start with some really wicked smart chatbots. Here I am thinking Moveworks.com. (Yes, I am starting to partner with them) Moveworks is an awesome technology built over the last 5 years by some very smart folks with experience at Google & Facebook. Moveworks gets it right. They understand how to interact with the customer, they know where they fit, they are easy to implement in a matter of weeks, you know what you should expect in service to the client before you ever even deploy and they help you make it your own. This little chatbot is resolving 30% + of the tickets at some of the most recognizable brands across the US in the first days post implementation. Talk about a quick win that everyone can see immediately. As an outsider, some of what I see as their special sauce is the ability for the bot to understand everyone, not just what it tried to pre-load as the five options for “reset password”. If you haven’t had a chance, go check out some of their videos on YouTube here.
Ok, so we have invested in a tool that almost immediately helped get our investment dollars back out – Employees happy they can get immediate service from an actual smart bot and not wait on IT for days, CFO happy with ROI, CIO happy with less times his top network engineer needs to show Bob how to print from Word…. you get the idea.
What’s next? – Here is where I think we are starting to work on the longer investment, more money more upside sorts of things… Next we need to go visit that RPA world and figure out which platform is best for us. Working with a partner (me, me, me) or finding a consultant to come in would be a good idea on this one. Regardless, you are most likely going to go find Automation Anywhere, UiPath or Blue Prism. Great, now what. Well this is where I think you work with the teams to identify easy wins again. Let’s not try to cure all the worlds diseases on day one. Easy win number one for most companies…. on-boarding and off-boarding. Not sexy but I promise HR, IT, Finance – they all have a bunch of annoying steps every time someone comes or goes and most of the time they are easy wins to show off that cool new RPA robot thing.
So what does RPA look like in this case? To start, these RPA organizations will already have a firm understanding of the key systems like Workday, SFDC, ADP and others to make that on-boarding/off-boarding process work. Many times they can sit over the shoulder (figuratively) using one of their discovery bots to watch your own unique process to fire John after he wrote so many terrible blogs that you just couldn’t take it any more. Once the discovery work is done, they can automate that process and now you have yourself your first scripted bot. (Maybe Ax Bot?). Next time you need to fire someone else, the process is just that much faster. Want to really show off? Let’s pull in that Moveworks chatbot you are leveraging and have the HR team just send your chatbot a quick message and say “please fire bob bozo”. From there Moveworks can kick off your RPA Ax Bot and its time to sit back and enjoy the show with the life of AI luxury.
Alright – so we have gotten the quick win with the chatbot, we are off and running with RPA and have our first business use case under our belt – time for the complementing step to help with rounding out our AI story. This is where I am thinking we turn to the integration & API platforms to really pour some gas on this fire. Tools like Dell’s Boomi, Mulesoft & others are making a splash in this space. With this effort we are really starting to tie all of our systems together regardless of if they are SaaS platforms or in a hybrid environment and helping create a more connected environment to help increase productivity and lower operational cost.
These projects, while at times complex, can really help deliver business value out of IT and move us to a place of helping enable AI now while building our overall journey. In the future, I see IT as does a peer of mine, Austin, who said we are going to look back at IT and think how silly it was that we had software that couldn’t learn from its self, make corrections and write its own future. While that may seem a bit like a movie to some, that future is near and I look forward to seeing its transformation as we all learn how to better support the business needs of today and tomorrow