AI for the Enterprise

Three lessons KPMG put to use with their Cognitive Lab

Share this post:

How do you innovate in an industry that is centuries old? You find new approaches to doing critical work in radically different ways.

KPMG LLP, one of the Big Four audit, tax, and advisory firms with a long history of leadership and innovation is doing just that. We’re bringing the power of AI to bear on our key service areas to help ensure we lead the way in the digital transformation of our industry and continue to bring value to our clients.

Central to this transformation is our cognitive lab, which opened to help us integrate new technologies like IBM Watson into both our own firm and into the work we do with clients. The lab brings our domain expertise to life in new formats, and creates powerful solutions and services around cloud, data, and AI that enhance and scale expertise.

What we’ve learned

1. We’ve found there are a few key aspects for success when approaching a client project: The right data, the right technology, and the domain expertise of our professionals

First, is data. In today’s world, much business decision-making has been characterized as a “big data” problem. For centuries, however, humans have been able to make decisions despite missing information. Data is a critical element to creating meaningful solutions – without it, our decisions would certainly lack an unbiased, comprehensive view. But it’s only one aspect.

Second, is technology. New learning systems, such as AI, are powerful technologies when applied to large data sets, and present us with unprecedented analysis and decision-making capabilities.

However, merely the combination of powerful artificial intelligence technology and large data sets alone cannot be the recipe for implementing a solution.

Third, is human expertise and domain knowledge. Intrinsic intellect has often allowed humans to make decisions in the absence of data.

Our ability to tap into that expertise – along with data and technology – is what allows us to create truly compelling enterprise AI solutions.

KPMG’s professionals bring decades worth of knowledge and the deep domain expertise necessary to creating the most meaningful solutions. This is the crux of the human/machine relationship – domain knowledge that helps bring new or seemingly irrelevant data points into context, which in turn brings insight to our professionals and leads to better-informed decision-making.

The KPMG Cognitive Lab is developing new ways to digitally capture domain knowledge and combine it with data and technology to create innovative assets that enable digital transformation of the enterprise.

2. Digital transformation must enable tangible outcomes

Our clients focus on tangible business outcomes. They want solutions to their business problems, and they rely on us as trusted advisors and domain experts to bring what it takes to the table.

At KPMG, we focus on outcome-driven conversations, as opposed to project-specific pilots, proofs of concept, and technology discussions. When we have conversations with clients about incorporating these kind of technologies, they often start with a simple, yet critical, question: “Can I make the technology work in my business context?” At our Cognitive Lab, in addition to leading scientists who are working with the latest technologies, we have invested in a small team of business strategists who are developing methodologies and frameworks that help our business clients answer the fundamental question of, “should we do this,” before taking on proof-of-concepts to answer the question of, “can we do it?”

When we go talk to clients about incorporating smart machines and AI into the solutions we’re crafting, the first thing they look for from a business-outcome perspective is hitting the key metrics they use to measure their business – increases in productivity, reduction in cost, higher revenues, better customer care, and so on. That hasn’t fundamentally changed.

This approach has helped our clients and stakeholders understand the digital transformation journey enabled by AI technologies in a more holistic fashion – how it empowers employees, what workflows have to be re-engineered and re-imagined, what cultural and organizational change elements have to be addressed in addition to data and technology elements. Planning for success beyond simply a technology implementation is key. Why waste a transformation opportunity?

3. It’s critical to rethink knowledge capture systems

KPMG is a knowledge and intellectual property company. Our IP and knowledge is rendered to the market through our experienced professionals. Even in a knowledge-driven company like ours, the way we capture and store knowledge is not designed for the smart machine era.

We must approach knowledge-capture in a way that allows us to impart it to this new class of machines, so that they can use it to reason similarly to the human mind. For example, the need to optimize document sizes for easy file storage and transfer has resulted in most documents created in Word or Powerpoint being converted to PDFs. We are realizing that the conversion process removes a lot of valuable metadata in order to optimize the file size. Smart machines care deeply about that metadata – it helps them ingest these documents and make sense of them faster.

As we move forward, we’re reevaluating our knowledge capture systems, which are designed for human consumption, and adapting them for the smart machine era. As you look at your business, consider what you would change knowing that machines will also consume the data. How would you capture knowledge differently?

What’s next

KPMG’s Cognitive Lab is focused on shortening cycles of training and teaching these smart machines to assist and augment human decision-making. We are creating new methods and approaches that combine domain knowledge, data sets, and technology to achieve this acceleration – one use-case at a time, until we have our own workbench to accelerate our digital transformation. Our technology alliance ecosystem, including IBM Watson, provides greater understanding and insights of technologies, enabling us to remain technology agnostic and help ensure that all technology fits our clients’ needs. This underpins how clients view us in our trusted advisor role with them. It’s part of our DNA – and is practically mandatory in our field.

What’s next for us has a great deal to do with empowering our professionals and people to deliver value to our clients. At a recent meeting, one of our senior managers said she welcomes the opportunity to use the technologies we’re building in the lab. The reason? It makes her feel that she was delivering value to our clients unlike ever before in her career at KPMG, instead of being consumed with “ticking and tieing,” as they call it in the accounting profession.

That’s exactly the the kind of digital transformation KPMG is investing in.

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

See how KPMG is driving innovation and empowering its employees through the use of AI

Principal, Intelligent Automation, Cognitive & AI, KPMG Innovation and Enterprise Solutions

More stories
June 20, 2018

Watson and Salesforce – Embedded AI solutions for your customer service and sales needs

Keeping customers happy, and maintaining a streamlined customer service process is more important than ever. AI-solutions have proven extremely valuable in mitigating the pitfalls of customer-service, especially when a business is using Salesforce to manage their customer relationships.

Continue reading

June 15, 2018

IBM Watson Studio named winner for best innovation in deep learning

We're thrilled to share that IBM Watson Studio was named winner for the "Best innovation in Deep Learning" at the world-renowned, independently-judged AIconics awards in London this week. IBM won the award for "being a pioneer in deep learning, helping to drive machine learning applications towards the true potential of AI."

Continue reading

June 8, 2018

How to prepare and clean data through quick operations, data profiles and visualization

IBM Watson Studio and Watson Knowledge Catalog include Data Refinery for self-service data preparation. Data Refinery puts data pre-processing and feature engineering in the hands of data scientists, enabling faster data insights.

Continue reading