“You don’t want to deploy tons of bots, because you’re going to automate bad processes. It’s important to be strategic about which bots you build,” Mr. Tysen told the Wall Street Journal.
Intuit’s strategy, he explained, started with a small team of just a handful of people who could focus on perfecting a single automation project. That “center of excellence” then enlisted the company’s finance and human resources teams to help build Intuit’s first bot, after which other divisions were enabled to build their own automated processes. Including divisions beyond IT in the development process is crucial, Mr. Tysen explained, to ensure that IT understands each division’s needs. All involved divisions should have a stake in the bot’s success.
The team members’ ability to parse out which parts of a specific process should be automated will drive success. Rather than re-engineering an entire process, he said, it may make more sense to integrate AI into only the most mundane and repetitive parts of it.
“It’s about IT creating the capability for the company and stewarding, teaching and igniting the passion for different teams, and then figuring out what makes sense for them,” Mr. Tysen said.
More articles about AI:
Viewpoint: AI holds promise, but won’t solve medicine’s biggest problems
Yale New Haven tests driverless hospital shuttles
Harvard Medical School’s AI screening system could produce more accurate cancer treatment programs
