Join Infinia ML’s Ongoing Conversation about AI
“The times where I see AI projects go well is when they focus on the ‘why.’
“Why do we need to do this? Why do we need to fix this?
“Is it because of COVID? Is it because of everyone working from home? Is there a corporate objective to reduce vendor costs?
“What is the why? And then the AI is just how.”
– Jerry Ting, co-founder and CEO, Evisort
It’s time for some good old-fashioned bedrock advice for putting AI to work for your business.
It’s the kind of advice that seems intuitive but we all need to be reminded about, like “Live by the Golden Rule” or “No cheese after midnight.”
In this case, the tips come from my conversation with Jerry Ting, the co-founder and CEO of Evisort, which uses AI for processing contracts.
Here are three lessons for any leader putting AI into practice.
1. Aim for Business Impact
See I told you these were intuitive. But people really need to get this!
“I think there’s no value business-wise in applying AI just to apply AI,” says Jerry. “It means nothing.”
“What business objective or stakeholder value are you generating by just having adopted a technology?”
Jerry says some companies get off track when they focus on technology rather than results.
“But for me,” he says, “is it AI, is it ML, or is it deep learning? It doesn’t really matter what the actual underlying technology is.
“If you can’t reduce costs or generate more revenue, you can have the best technology in the world, but from a business perspective, who cares?”
2. Pick a Specific Business Problem.
Some large companies, says Jerry, are trying to build AI for everyone to use.
“But if it works for everybody,” he points out, “that means it works very well, in my opinion, for nobody.”
“And so what you have to do is if you’re looking at a problem space, you have to be deeply verticalized, deeply focused.”
Jerry’s company Evisort, for instance, is focused on applying AI to contracts.
“Maybe in the future we’ll do other document types, but it has to be related to the deal,” he says. “We’re not going to go out and do just any type of document.”
“And that’s a difference in terms of accuracy, because we’re solving for a known problem, versus if you’re building AI for everybody, your problem is inherently not solvable.”
3. Rethink How People Work
“For me,” says Jerry, “it’s about augmenting and giving time back to the people that use us. It’s not about replacing the actual people.”
“We don’t know what the ideal use of a human in the loop with a partially automated process looks like, but what my gut tells me is that that human is going to be allowed to do what they actually want to do, what they went to school for, what they want to apprenticeship for.”
“I started my company having been a lawyer because I found that 80% of my time was used going through documents and looking for the same things over and over again.”
“And honestly, that’s why I stopped being a lawyer is that that’s not why I went to law school. I went to law school so that I could be consultative, so I can be strategic, so I can be a business partner for my customers.”
Jerry says that’s still what he’s about today — “but I think how we get there is not work harder and work more nights and weekends, it’s actually understanding and designing the right software with artificial intelligence, so that we have more time to actually be strategic and be humans.”
So what’s next in Jerry’s world of AI contract processing?
“I think tomorrow we’ll actually negotiate your contracts for you,” he says.
“We can actually take a contract that’s been negotiated and redlined, it comes into your inbox, we can tell you, ‘Hey, these areas are ok to accept, these areas you should push back on. And not only you should push back on, but here’s how you should push back.’”
“And I can imagine a world where that negotiation process is automated. AI on both sides, doing that negotiation back and forth, and then up to the humans at the end to accept that and approve the work.”
“But the back and forth of going via email, I think that’s going to be a thing of the past.”
In fact, Jerry says Evisort is doing this with two clients already. The tomorrow he’s thinking about is already here today.
Thank you for reading! Please subscribe, like, share, comment — give the algorithms what they want!