AiThority Interview with Gary Kotovets, Chief Data and Analytics Officer at Dun & Bradstreet

Welcome to the Technology Interview Series. Please tell us about your role at Dun & Bradstreet.

I am the Chief Data and Analytics Officer at Dun & Bradstreet, a leading global provider of business decisioning data and analytics. In my role, I lead the company’s data and analytics strategy globally and am responsible for enterprise data governance and exploring new digital and alternative data and analytics opportunities.

In my role as CDO, I like to start from the end from the client’s point of view first then work backwards into the guts of the company to understand/change/design/architect its data and analytics as a product to fit clients’ needs. This is why, for instance, recently we announced the launch of D&B.AI™ Labs. Many of our customers are looking for ways to capitalize on the opportunity offered by generative AI and large-language models. We wanted to create a cutting-edge environment where we could accelerate our customers’ innovation efforts and co-innovate with them in a secure and responsible way.

Dun & Bradstreet is close to a 200-year-old company with multi-decade relationships and helping more than 240,000 clients around the world, including 93% of the Fortune 500, tackle some of the world’s biggest and emerging challenges. Dun & Bradstreet holds information on over 500 million business entities through our D-U-N-S® Number that has become the world’s most recognized corporate “fingerprint” or identifier for connecting business entities, linking relationships, and validating financial health. Our data solutions are helping companies not only grow, but also help measure and mitigate business risks with increasing complexity and importance.

Recommended: AiThority Interview with Abhishek Shrivastava, VP of Product at LinkedIn

What has changed in the business intelligence and data analytics landscape since the launch of ChatGPT?

Generative AI (GenAI) and Large Language Models (LLMs) are unquestionably a powerful tool, that with time will become even more beneficial for businesses. As powerful as these models are, the onus of thinking is still on people.  Businesses must clearly understand when the use of generative AI and LLMs is appropriate or not and have a comprehensive GenAI/LLM strategy. For instance, do they use open-source models or buy proprietary models, do they implement models off-the-shelf or do additional training and fine-tuning inhouse. The latest and most powerful model may not necessarily the best one for each use case.

The data angle has become even more important than ever before. The AI models are only as good as the data that goes into them. Historical, clean, structured, organized, validated trusted data will yield significantly better results and will allow businesses to incorporate these capabilities into their operations and decision-making processes.  In a world where LLMs are trained on pretty much the entirety of publicly available digital data –mainly uncontrolled publicly available data from the web – the value of trusted datasets that are underpinned by validated, historical and proprietary data grows exponentially. It is thus more important than ever to get your data strategy right.

Finally, tighter governance is essential due to numerous emerging types of cybersecurity risks. There is greater accessibility for companies to utilize AI models, such as ChatGPT, to drive improvements with better efficiency and productivity in business operations. Businesses are getting savvier in the utilization of AI models but need to be mindful of the many risks associated with it.

While much about LLMs remains to be understood, the models are the new normal. Businesses that ignore these developments risk falling behind. But with the democratization of this technology comes increased responsibilities which businesses need to be consider. It’s an exciting time to be in the data and analytics field, with the pace of change accelerating and new opportunities arising. It will be fascinating to see how this continues to develop and how businesses embrace it.

Could you highlight Dun & Bradstreet’s approach to innovating with generative AI tools?

The voice of the customer is our driving force for making technological and operational decisions about how we were going to put that data together and serve it to our customers in a more meaningful way. Once you establish that, you improve the engine and serve your customers with the best coverage and the best analytics, you then have the opportunity to improve and tweak that engine to continuously optimize those capabilities to serve the customer.

GenAI and LLM have the potential to transform how we collect, analyze, and use that data and the insights we can generate for our customers. This is why we have been investing in these technologies over the last several years and have already had several solutions deployed that use large language models.

Our customer-focused approach leads all our efforts in creating a safe environment for clients to fulfil their use cases that deliver value within the AI realm. Dun & Bradstreet has a tremendous amount of valuable proprietary data, so our course of action is to understand the challenges that our clients have and rapidly deploy prototypes that address those needs.

Recommended: AiThority Interview with Rob Walker, President of Global Growth Markets at Cognizant

Please help us understand the application of D&B AI Labs.

D&B.AI Labs is a secure environment where we will work side-by-side with our customers by using our proprietary data and analytics and their information to formulate solutions in real-time, build prototypes and rapidly deploy solutions by applying Gen AI, LLM, and other technologies. It is responsible innovation that will identify the pain points of customers and develop solutions tailored to their specific needs. Dun & Bradstreet has gone through a dramatic transformation driving a culture of innovation and making significant investments in technology, data, and analytics, including adding 64% more analytics solutions, evolving its scores and indices to leverage AI, LLM and ML capabilities.

D&B.AI Labs comprises a team of seasoned data scientists, data engineers and solution specialists with extensive innovation experience, as well as deep expertise in AI, LLM, ML, and advanced business analytics who have a track record of successfully implementing solutions for a variety of use cases. Companies need the insight, direction and confidence that only comes with clean, actionable data as the quality of data you put in directly impacts the quality of the data you get out.

Our products and services are underpinned by validated, structured, historical and proprietary data, which allows us to deliver reliable and interpretable AI-created results that drive our clients’ most critical business decisions.

A lot of heated arguments are going around when it comes to discussing the accountability, biases, ethics, and trustworthiness of AI models. Could you please tell us how you built a responsible AI model for Dun & Bradstreet?

Today, more than ever, knowing what data to trust and having the right tools to make that data actionable is what drives a competitive edge. However, users of AI models need to understand all the risks involved including six major factors: quality of the generated insights, ethical, legal and regulatory risks, security and cost/ROI.

Dun & Bradstreet established its AI framework within these six parameters and has a cross-functional team comprised of experienced legal, compliance, and data scientist team members that direct and guide the company’s use of AI models. The company has established trustworthy relationships over its nearly 200-year history and has formed many strategic partnerships with foundational model providers to ensure we are responsibly applying GenAI to our data and working with clients to test and help train as they use our data repository.

Dun & Bradstreet has the right tech stack, right talent and deep understanding of which models work best for which use cases, and we have an established robust data infrastructure that makes use of LLMs for custom business uses quickly and efficiently.

What kind of training is required to operate D&B.AI Labs? Would you be providing additional knowledge management tools to simplify the ease of adoption and use for this new AI platform?

Yes, training is critical and will be provided to help clients with the understanding how to, for example, build pipelines and do prompt-engineering in our environment to get to the best possible outputs. Also, the reason we like working with our customers and co-innovating with them is that we want to help them accelerate their innovation efforts– we want to understand the problem they are trying to solve, point them to the right datasets in the Lab environment, agree on the most relevant approach and prototype a solution together. Our customers have some of the most sophisticated data science teams in the world, but we want to flatten the learning curve for them when it comes to understanding the multitude of our proprietary datasets and articulating the right solutioning approach.  And last but not least, as we go through the process of working with customers, we’re hoping to educate them on the new types of risks and potential governance concerns and on how they can deploy these technologies in a secure and responsible fashion.

Recommended: AiThority Interview with Michael Schmidt, Chief Technology Officer at DataRobot

Your predictions on the future of AI in the customer service industry that requires highly optimized and CX-driven workflow automation.

The use of AI in the customer service industry will continue to help enhance and augment existing service offerings, helping to improve productivity and the amplification of skilled workers. AI models are sometimes referred to as co-pilots, meaning they will serve to help and strengthen existing services – not replace. The best utilization of AI is not to replace human interaction but to enhance human interaction and decrease the friction in the customer experience. Human elements are still needed to interface and validate answers with AI models to ensure the output is correct. AI can help improve customers’ and users’ experiences and help gather pertinent data on customer behavior patterns and help predict future roadblocks.

Customers want more personalized service and experience that can be gained by AI; however, AI needs to be injected into well-established customer service procedures and processes. Businesses need to be careful to not underwhelm the customer by removing the human touch as human workers are still the most important assets of a company.

Thank you, Gary ! That was fun and we hope to see you back on AiThority.com soon.