How a top-tier professional services firm decodes tough tax law to get millions back for their clients

For KPMG and Watson, confidence is in the details

How a top-tier professional services firm decodes tough tax law to get millions back for their clients

For KPMG and Watson, confidence is in the details

Client:  KPMG
Industry:  Tax and audit


When dealing with taxes for major corporations, “complex” doesn’t begin to describe the task.

That’s why businesses seek the help of KPMG LLP, a Big Four audit, tax, and advisory firm, especially when it comes to R&D tax relief. Tax professionals and engineers in R&D departments spend thousands of hours manually reviewing materials to discern the appropriate evidence, and missing a detail could mean missing out on millions. “The government stands to give back $148 billion in the next 10 years, but getting those rewards isn’t easy. It’s hard to build up detailed evidence to qualify,” said Brad Brown, Chief Innovation Officer, Tax, KPMG LLP.

So, KPMG turned to Watson to help them more effectively find R&D tax relief for their clients.

How Watson learned in 5 steps:

  • Watson was trained

    Watson absorbed over 600,000 pages of documentation, from reports to correspondence.

  • Watson was tested

    The machine learning model was continuously updated to be able to analyze a higher volume of records.

  • Watson was launched

    Over 80% of employees adopted Watson for their day-to-day work.

  • Watson got results

    Employees used to spend 80% of their time researching problems and 20% fixing it; Watson has reversed that.

  • Watson keeps learning

    The employees are encouraged to provide feedback, be it as a new, intermediate or experienced user.

Why Watson?

“We wanted to transform the industry,” said Brown. But embracing change meant finding a partner they trusted. “After researching all our options, we realized IBM was the technology provider we wanted to work with,” he said.

“KPMG needed a partner that could bring in the latest AI technology and align with our vision and innovation strategy. That’s where IBM came in,” said Steve Rainey, KPMG, Chief Innovation Officer, Tax. “IBM has had a longstanding strategic alliance with KPMG,” said Hill. “Working together was how issues were solved early on. It was a big part of our success.”

Robust, yes. But is AI ready for taxes?

“Starting out we knew this would be challenging,” said Brown. AI had proved helpful in other industries, but tax law is lengthy, tricky to navigate, and oftentimes difficult to decode. “Watson had to meet the expectations of our clients and people to make them want to leverage it,” said Brown. “We had to have meaningful training for Watson to build a worthwhile knowledge base.”

Vinodh Swaminathan

KPMG, Principal

Todd Mazzeo

KPMG, Managing Director

“Watson helps me deliver better results in less time. That’s a huge advantage.”

If AI was an answer, it had a lot of learning to do.

“Watson isn't a PhD grad out of the gate. It starts off as a kindergartner and works its way up,” said KPMG’s Todd Mazzeo. But with training, Watson got better at helping tax professionals determine tax relief eligibility with confidence. “Watson becomes smarter with every new document we give it,” he said. And after integrating Watson into their workflows, KPMG’s tax advisors could more quickly examine massive amounts of information, like loan assets and nuanced tax interpretations, to help them offer clients richer, more detailed recommendations.

Related content

Business professionals must be seen as reliable advisors, able to provide recommendations and advice their customers can rely on. Learn how you can make informed decisions with AI.

Read the smart paper →

APIs used include:

Watson Discovery →

Watson Knowledge Studio →

Watson Natural Language Understanding →

The Watson advantage

Using 10,634 documents, Watson was trained on five different R&D models by 13 tax professionals. After training, Watson recommended the correct tax treatment about three out of four times. “This means confidence in numbers for our tax professionals, because now we have an incredible audit trail of how to get to these conclusions,” Hill said.

“Watson helps us be more productive,” Hill said. Employees are now able to analyze data and find deep insights more rapidly. Brown calls it an “energizing” experience. “The change Watson has brought internally is making a difference in people’s job satisfaction,” he said. “People are happy and motivated to come in and do something exciting.”

Here's what makes Watson AI for smarter business:

  • Watson learns from less data

    Watson’s algorithms are specially tuned so that your businesses can rapidly achieve accurate results from smaller, private data sets.

  • Watson reimagines your workflows

    Because Watson can be embedded within the many platforms and applications you already use, you can improve most aspects of how your business works.

  • Watson protects your insights

    You need to own your data, your algorithms, and your IP — with Watson. you can. The insights gained from your data are the foundation for your competitive advantage, and they stay yours.

Sri Varanasi, KPMG, Senior Manager

“It was exciting to see Watson’s results match ours.”

Sri Varanasi
KPMG, Senior Manager

Steve Rainey, KPMG, Chief Innovation Officer, Tax

“We can put Watson in the hands of our experts and get effective results.”

Steve Rainey
KPMG, Chief Innovation Officer, Tax

How Watson learned, in five easy steps:

  • 1. Watson was trained

    Watson was trained on five different R&D models by a tax team of 13 with more than 10,000 documents.

  • 2. Watson was tested

    End users tested Watson early and often to develop a system that worked the way they did.

  • 3. Watson was launched

    Watson began helping with research credit services and now helps five more service lines transform their offerings.

  • 4. Watson got results

    Watson recommended the correct tax treatment about three out of four times.

  • 5. Watson keeps learning

    Watson learns from each interaction and gets better at generating more detailed results every time.

The future for KPMG and Watson

“The way we’re applying the technology is building excitement,” said Kay Sugarbaker, KPMG LLP Project Director. The project’s success is building interest around the office, with more departments excited to bring Watson on board. Vinodh Swaminathan, Principal at KPMG LLP, is ready for what’s to come. “With Watson, we’re able to offer our clients products and services we couldn’t imagine before,” he said. As Brown put it, “that’s a big homerun for us.”

Some or all of the services described herein may not be permissible for KPMG audit clients and their affiliates or related entities.

For developers & IT

Build innovative cognitive apps with Watson on IBM Cloud.

Try IBM Cloud platform for free →

Connect with us

Talk to us about your use case to learn how Watson can help solve your most complex business challenges.

Connect with a Watson rep →