For data scientists at some of the country's biggest companies, artificial intelligence is all about the details.
At the Activate 2019 conference, hosted by Lucidworks in D.C., four companies participated in a new competition evaluating blossoming AI projects in e-commerce and digital enterprise management.
The Connector Awards celebrated novel uses of AI – in both small and large-scale projects – from companies including Uber, PwC, Lenovo and Nordstrom Rack/Hautelook.
Judges for the competition included representatives of leading technology analyst firms, Sheryl Kingstone from 451 Research and Mike Gualtieri from Forrester, as well as DC Inno’s Kieran McQuilkin. The contest determined two winners: One for "Digital Commerce," or customer-facing search and AI innovation, and the other for "Digital Workplace," based on intracompany platforms.
First to present was Marc Desormeau, head of data optimization at Lenovo, which sells hardware and other supplies for offices, schools and companies. As a seller of products ranging from personal computers to data center equipment, the company needed to accommodate both one-off consumer purchases and high-price enterprise sales.
Using AI as part of its e-commerce search engine immediately impacted the company, significantly raising click-through rates and revenue generated from search within months. Lenovo is creating different product search paths for different users, and using the data to better categorize its inventory. In addition, Desormeau's team created a dashboard where suppliers can view and organize their own search metrics from the Lenovo site, freeing up time for client-facing staff and increasing transparency for those clients.
For Pankaj Andhale, a data scientist at Nordstrom subsidiaries Nordstromrack.com and Hautelook, a new platform using AI was able to combine two search engines that previously operated separately – one for Nordstrom Rack's online store and one supporting Hautelook's physical stores. Similarly to Lenovo, the search engine results are shared with merchandising teams to create a feedback loop and constantly update its algorithms.
Viren Patel, who spoke at the conference, is building PwC's enterprise search platform using AI. Departments across the large company are using its new search engine to find technical and HR documentation, thereby spending less time looking for the right person in a sea of colleagues. The platform has been so successful early on that PwC is beginning to sell "enterprise search-as-a-service," where other companies will use its tracking system for institutional knowledge and resources.
At Uber, AI-enhanced search functions are used for more specific, but more complex purposes than at PwC. Atousa Duprat, software engineer at the rideshare giant, is building an internal search engine focused on finding and organizing complex documents across the enterprise. By aggregating employees' search queries and navigating through a maze of hierarchies, Uber's AI models boost relevant files and documentation to the top of the list.
All four initiatives have showed promise at their respective companies, but in the end, the competition could have just two winners: Nordstrom Rack/Hautelook for Digital Commerce and PwC for Digital Workplace.
This is the first of many Connector Awards competitions, according to Lucidworks, which plans to host its second effort at next year's Activate conference.
