Unleashing Intelligence: Leveraging AI to derive Insights at Scale

Generative AI has proven to be an impressive tool for industries in generating insights from a large volume of data. Parmjeet Virdi Director Data Analytics at Publicis Sapient, reveals while speaking at AIM Media House’s Data Engineering Summit (DES) 2024 that Publicis Sapient has been leveraging generative AI to derive insights from customers data and helping them make better business decisions.

To make her point, Virdi referred to an incident of an analyst reviewing a lengthy report by Harvard Business Review. She said an analyst overlooked crucial details in a lengthy report, missing the competitor's expansion strategy. 

However,AI, unlike humans, connected scattered information, revealing the competitor's plan to penetrate a low-cost market by leveraging specific product manufacturing and later market expansion, enhancing analytical insight.

The initial disruption caused by the internet paved the way for subsequent technological advancements, including the evolution of web, mobile, IoT, and connected systems. So when these technologies are evolving, at the same time, the customers are also evolving,” Virdi said.

Along with Aashima Kumar, senior manager, data analytics at Publicis Sapient, discusses how Publicis Sapient has harnessed the power of generative AI, Machine Learning models, and advanced Data Analytics techniques to unearth hidden patterns, trends, correlations, and influential features within extensive data repositories.

Deriving insights from data 

Moreover, Kumar pointed out that Publicis Sapient built a solution for a large automobile company that helped them drive sales by better understanding customer data. 

“So when any potential customer comes to their website or mobile app, they see a very fancy-looking configurator that gives the customer the option of selecting the model, grade, color, engine, upholstery, and accessories, plus all the subscriptions for the connected cars.

“Now that so much of data coming out of configurator and talking about this data, it actually has the power for us to understand the needs of the customer, at the same time provide back to the business a way that they can understand what kind of combinations are more profitable for them and in more demand.

“We utilised conjoint analysis and Monte Carlo Markov analysis to identify coefficients for customer choices, generating insights visualised through a utility dashboard,” Kumar said.

This internal tool aided product planning and experience teams by revealing utility scores, indicating component attractiveness and profitability trade-offs, and enhancing decision-making across multiple business levels, including product combinations and profitability assessment.

Leveraging generative AI 

Marriot, one of the world's largest hotel brands, had a traditional search engine, Kumar pointed out. Marriott's traditional search process needs users to input destinations, dates, and preferences, among other things. However, this method leaves much to be desired in aiding decision-making, offering limited assistance.

“We transitioned their traditional search to a generative AI-powered search, enabling users to input specific needs in natural language rather than filters. The generative AI model then extracted these queries into attributes and identified relevant accommodations.

“We scaled it to the next level to make it a conversational AI. Using the same power that we were able to engage with Gen AI, we made sure that it's not just giving you the options of what you're looking for. But why can't we recommend you more options and then you make the choices that you want,” Kumar said.