In accounting, a little artificial flavor goes a long way

Elementary, dear Watson: Applying IBM Watson's artificial intelligence to audits and lease reviews frees up human minds for bigger things, according to EisnerAmper Managing Partner Jay Weinstein.
By GREGORY ZELLER //

Not all of artificial intelligence is working to infiltrate national defense systems and annihilate humanity in an apocalyptic firestorm. Some AI just wants to get you a better tax return.

That’s the human thinking, at least, as artificial thinking spreads throughout the matrix at New York City-based accounting giant EisnerAmper. The firm has already plugged AI protocols into various corporate offices, and will bring them soon to EisnerAmper’s thriving Syosset satellite.

And not just any AI but the much-ballyhooed IBM Watson, standard-bearer for the commercial (and beyond) potential of artificial thought.

Through a professional partnership first suggested in 2017 and officially launched in 2018, EisnerAmper and IBM are training Watson to think long and hard about audits, contracts and other staples of corporate conduct – a value-added effort already paying off in multiple practice areas, according to EisnerAmper Managing Partner Jay Weinstein.

Based in Central New Jersey, Weinstein is the firm’s managing partner of markets and industries, putting him in charge of both EisnerAmper’s market groups and its IBM Watson program.

“My job is to lead our growth initiatives,” he noted. “And we view AI as a tremendous area of growth – another way to help our clients more efficiently manage certain areas of their businesses, and help us more efficiently audit them.”

Jay Weinstein: Organic growth for artificial intelligence.

Weinstein – who earned bachelor’s and MBA degrees in economics from Temple University and a master’s degree in taxation from Villanova University – is not inherently a computer guy. But he’s “in lockstep” with EisnerAmper Chief Technology Officer Amir Segev, with whom he co-leads the Watson effort, and he’s benefitted greatly from his exposure to IBM’s Watson teams.

“They were very easy to work with,” he told Innovate LI. “We’ve had a very positive experience with them.”

Now, the mission is to bring the positivity to the people, or at least EisnerAmper’s international client base.

It’s already been a hefty investment. “We spent a whole bunch on it,” Weinstein noted, and as the protocols spread throughout the international company, new licensing fees will rise.

But the training costs, sweat equity and other expenses are quickly proving worthwhile, according to the MP.

“It’s expensive and it’s time-consuming,” he said. “You get your little piece of a server at IBM and you train your own ‘instance’ of Watson, starting from scratch.

“But you educate it the way you want to educate it,” Weinstein added. “And in the end, it’s beneficial.

“There’s a return there.”

The return, so far, has been felt mostly in the auditing arena. EisnerAmper initially teamed with IBM to apply Watson’s AI abilities to a new “smart auditing” procedure, designed for more effective and efficient auditing of pension plans, for instance.

But over the last month, EisnerAmper has started incorporating AI protocols into non-audit services – using it to review contracts and leases, for instance, and also applying Watson to tax returns.

“Watson will read a tax return and see certain things and ask questions,” Weinstein said. “‘I see you are a manufacturer. Did you take the proper manufacturer tax credits?’

“And it will learn,” he added. “That’s the beauty of Watson. Next time, it won’t ask that question – it will ask different questions.”

EisnerAmper has been slow-and-steady regarding AI introduction to its workflow. Weinstein noted specific efforts to engage company thought leaders before new tech is incorporated, to ensure changes are met with open hearts and open minds.

“When we ask our folks to adopt new technologies, we always ask about their mindset to make sure our colleagues are open to change,” the managing partner said. “Once they are, then the adoption of technology like this is relatively simple.”

The idea is not to replace human minds, he added, but to “free them up to do more interesting and challenging things.”

“They can analyze the data that comes out of Watson, instead of spending time reading through a 30-page lease,” Weinstein said. “That’s much more interesting and fun.”

The Long Island team is absolutely open to the change, and the EisnerAmper industries chief expects Watson to be on board in Syosset sometime this summer. Outside IBM consultants will spend a matter of days instructing Long Island staffers on playing nice with their new AI associate; those staffers will essentially create their own version of Watson to meet their specific needs.

“The chief beneficiaries on Long Island will be our audit colleagues,” Weinstein noted. “And we’re starting to scope out new tax projects, which is also good for Long Island, which has a really strong tax group.

“The point is to have the right people train Watson, so it asks the right questions,” he added. “And we have subject-matter experts around the entire firm.”