Business Process Management
Invoicing should be easy. A contractor performs work for your company, submits an invoice and is paid by Accounts Payable within 10 days, right? Not so much. Accounts Payable often gets backed up. Sometimes it takes months to get contractors paid. And while those contractors are becoming increasingly irate about not getting their money, your organization is losing money.
It's a truism often tossed around among software engineers: the faster you can find a bug, the cheaper it is to fix it. Lately, we've been publishing a series about DevOps — the combination of development and operations teams, de-siloing IT, embracing automation and extending agile development practices past development and across the applications lifecycle. Speed is a huge benefit of DevOps — when workflows are automated, code is released faster.
A couple of weeks ago, I wrote a post about DevOps and why all industries need to implement it. But implementing DevOps is not as simple as combining your Development and Operations teams. Any organization that wants the full benefits of DevOps also has to automate things as much as possible: testing, compiling, deployment, even creation of your servers.
It's a customer experience fact: everyone who calls customer service wants to talk to a human being. They mash 0 to avoid self-service. They tweet at the company when they're on hold. They try to contact someone, anyone who can answer their question.
Have you ever been part of a large corporate IT project that got all kinds of attention during implementation and launch, only to die on the vine a few months later? Many companies don't need to look past their own web site. Everyone pays attention to it in the design phase. Everyone wants a say in how the home page looks. Every department wants a chunk of home page real estate. And no one will turn down a fist-bump or beer when it launches.
There is an unplanned outage and your company's most important application has gone down. It's taking forever to come online again. When you ask your IT department what's going on and when your application is going to be up and running, all you seem to hear is blame. The development team is mad at the operations team. The operations people are mad at the development people. Everything seems to be someone else's fault.
Call centers are broken. And unless someone does something about it, contact centers will continue to be the villains of the customer service world, funneling organizations' cash into a black hole, and sucking the life out of customers who are waiting on hold.
You've got database drama. You've tuned and tuned, but your queries are still taking forever, and when they come back, they're not returning great results. You have to do something, but what? Denormalize? Change your schemas? Buy a bigger server? If you've already done everything you can possibly do already, you know none of those steps is the right solution. The problem may be that you're using the wrong tool for the job.
When artificial intelligence is discussed, people tend to get a little breathless about its possibilities. And why not? The prospect of developing human-like intelligence and incorporating that intelligence into everything from cars to the apps we use at work is exciting. Humans have been dreaming about AI for decades, and we’re just getting to a place where AI is catching up to our fantasies about it. What might we be able to achieve when machines are doing some of our thinking for us? Here’s the thing though; machines have been doing some of our thinking for us — sorting through data and making predictions — for decades, but it’s so commonplace that no one gets excited about it.