US-based Dun & Bradstreet provides commercial data, analytics, and insights for businesses. The company started its Indian operations in 1995 and has offices in Delhi, Chennai, Bangalore, Kolkata, Hyderabad and Ahmedabad.
Analytics India Magazine spoke to Avinash Gupta, Managing Director & CEO – India, Dun & Bradstreet, to understand its data and analytics offerings in India. An alumnus of IIT Varanasi and Tulane University, Avinash comes with more than 30 years of experience.
“Our AI experience lies in the way we gather, curate, and validate our data, develop, and test our own models, and use AI in all our products and solutions. Additionally, we leverage AI methods and techniques against customer-specific challenges to develop custom predictive models and data management,” said Avinash.
AIM: Tell us about Dun & Bradstreet’s strategy for India
Avinash: We believe that businesses in India can benefit immensely by leveraging data and analytics and use embedded AI and analytics in the business decision process. We have supported Indian businesses in areas spanning finance and risk, sales and marketing and third-party risk and compliance.
We have been researching, developing, and employing AI and analytics to elevate the value of our insights. It is the key enabler to running our business, starting with the methods we use to gather, curate, and validate our data and carrying through the development and training of our more than thirty predictive scores and indicators. The opportunities are immense, and we are at the forefront of enabling AI and analytics adoption for our customers in India.
AIM: How do you take advantage of AI and ML?
Avinash: Leveraging D&B Data Cloud — the world’s most comprehensive commercial data, we enable transforming business data into predictive insights to drive operational optimisation. We use AI and ML to enable businesses to optimise risk, gain operational excellence, and drive profitable growth. We have gone the extra mile to offer these products as self-service platforms to our customers to use our commercial data and a pre-built library of machine learning models catering to business use cases. Some examples include sales and marketing analytics: data vision, advanced analytics: analytics studio, risk: finance analytics, compliance: OnBoard.
Our predictive risk models are critical to businesses, predicting risk-related outcomes including insolvency, failure to pay, fraudulent behaviour, and the impact of natural disasters. However, these same capabilities allow us to predict positive outcomes, predicting those businesses poised for economic recovery and growth and those firms targeting government programs. This allows businesses to make faster, more accurate decisions about whether to partner with an entity, reduce costly disruptions in the supply chain, reduce dependencies on critical suppliers headed for business deterioration, and save precious budget and resources on on-site visits/inspections. Dun & Bradstreet’s AI and analytics products and solutions for our customers in India include:
The machine learning methods offer greater precision of prediction, capturing the effect of subtle changes in data and granular segmentation. Human intelligence creates explainable results from ML-driven models. We are leveraging data to accelerate evidence-based decision making: As a data-inspired organisation, we know that descriptive and predictive models are only as good as the data that feeds them, so we consistently focus on the ability to leverage data in AI-enabled solutions for our customers.
AIM: How AI and data science help in building products and services?
Avinash: Dun & Bradstreet uses an AI & Analytics first approach to conceptualise any product and solution, including re-architecting our legacy products and embedding AI & Analytics-driven insights that our customers would need for business decisions.
The visual below demonstrates how Dun & Bradstreet integrates AI in developing advanced analytics products and in customer engagements where we build custom predictive insights. Each work block indicated with the AI icon is enabled and realised through our in-house AI capabilities.
At the same time, Dun & Bradstreet is highly focused on ethical principles in AI as part of its product strategy. Accordingly, we embed the following ethical considerations into our practices and methods.
- Comprehensive focus on permissible use, including all relevant regulations, and considering the cross-border transfer of information and data localisation requirements.
- Explainability for credit-facing scores and analytics, including what information was used and a treatment of potential bias (e.g. missingness of data).
- Privacy embedded in all solutions (e.g. compliance with GDPR and similar regulations globally).
- Provenance is also maintained, whereby methods and solutions are tied to traceable data sources and elements.
AIM: Give us a glimpse into D&B’s data science culture.
Avinash: We have significant talent and expertise in AI, predictive modelling, and cognitive analytics and a highly motivated team of AI SMEs, data scientists and ML Engineers to drive our products and customer engagements. We have built and groomed team members who aspire to make the best possible use of D&B’s commercial data to derive insights that have a far-reaching impact on the business decision process for our customers. And we continue to keep our people motivated through various HR initiatives to give their best to the organisation.
AIM: Can you share a few AI/ML/data science use cases?
Avinash: Few use cases of using AI and ML, including NLP, deep learning, network science and many more:
- Sales & Marketing: Use cases to score and prioritise prospect accounts using D&B commercial data and external data, including web data and other publicly available data based on intent, the likelihood of conversion and other metrics.
- Finance & Risk: Use cases including credit risk scorecards using ML for credit scoring, supplier risk modelling, loss modelling, collections prioritizations models and others that use traditional predictive methods as well as recent developments in the field of AI.
- Industry Categorisation: Using ML-based neural nets to look through the information about a company on its website and from other sources. We have models that leverage natural language processing (NLP), machine learning and semantic search to help identify and assign an industry classification code for a company based on what that company says on its website.
- Web Discovery: Automated web discovery tools to uncover new executive-level contacts from publicly available corporate websites, classify contacts in real-time, and allow our customers to search and find relevant contacts faster.
AIM: Tell us about D&B’s hiring process.
Avinash: At Dun & Bradstreet, we have created robust training programs to build in-house competencies while at the same time looking for talent to join us across all our products and platform development teams. We look for passion, a problem-solving mindset, and business understanding and the technical skills that we need to build our products and platforms.
We are continually looking for talent to help us differentiate in the marketplace. We have been hiring at a rapid pace across all key cities to build strong products and engineering teams and front-end roles in sales, marketing and customer support.
AIM: How does data science fit into your business model?
Avinash: We focus on business priority solutions. Dun & Bradstreet collaborates with businesses to solve their most challenging data and insight problems with AI-enabled solutions. Additionally, we aim for data that accelerate evidence-based decision making. Combining validated, rich data with AI makes insights and decisions faster, simpler, and more accurate. With the Dun & Bradstreet Data Cloud of more than 400 plus million unique businesses, our suite of powerful predictive indicators, and our machine learning entity resolution processes, businesses can elevate technology offerings with data-centric solutions.
AIM: What is your roadmap for AI and data science applications in 2021?
Avinash: We are constantly building, upgrading our platforms and applications to embed AI & ML use cases to enable our users to benefit from the predictive insights. Examples include launching a leading Sales and Marketing Rev Ops platform that combines a constantly evolving data landscape with the world-class company and contact identity resolution, predictive insight, workflow integration, and multichannel activation.
New areas involve using unstructured data, including text, web links, pictures etc., to extract unique business activity signals. A vital priority area includes scoring companies based on their ESG performance and identifying the risk associated with a large section of Indian businesses. Other priority areas include streaming data analysis and near real-time alert generation for financial risk monitoring systems.