Lucy® the Answer Engine® broadens her knowledge and scope with Azure

Solving an enterprise scavenger hunt

What was that analyst’s response to our 2022 annual report? How did the focus group react to our new product feature? Where can I find the latest employee expense report training? If you’ve ever scrambled to find corporate data, you can appreciate a solution that eliminates that frustrating hunt for information.

The founders of Lucy recognized that inefficient processes, employee turnover, and other corporate activities can make it difficult to share knowledge or insights—or worse, lead to data loss. As early adopters of artificial intelligence (AI), the founders saw an opportunity to build a solution that removed these barriers to search. They began work in 2015, using IBM Watson to develop a platform to solve the daily hunt for information. (“Lucy” is a nod to Lucinda Watson, granddaughter of IBM founder Thomas J. Watson.)

With its innovative approach to creating an “answer engine®” rather than a search engine, Lucy (the company) reinvented knowledge management. Lucy (the platform) uses natural language generation to read and synthesize millions of search results that “she” finds across an organization’s knowledge ecosystem. Lucy brings you to the exact answer you need—at the specific page, slide or time stamp it can be found. Users can also ask Lucy a question in a private chat or mention her in a group chat within their preferred business messaging platform, such as Microsoft Teams or Slack, to get an immediate and summarized answer. 

When enterprises 'hire' Lucy, they don’t need to move their data. Companies maintain full control over what assets Lucy can search, including granular, role-based permissions that they can change as needed. Assets are stored within secured internal systems, ensuring compliance with security and privacy requirements.  

The evolution of an answer engine

There have been several iterations of Lucy since 2015. Each customer’s solution has been tailored according to business needs using the latest technology. Before her latest version, Lucy could search Teams environments, SharePoint sites, SAP, third-party tools, and more, but there were unstructured sources such as videos that she couldn’t review.

With Lucy’s growing prowess in serving Fortune 500 companies, the amount of corporate data, both structured and unstructured, continues to increase exponentially. Lucy needed to evolve from her previous configuration to address this data proliferation. And Lucy needed a modern AI solution and cloud infrastructure it could tailor to its customers’ needs and that could evolve as customers—and corporate knowledge—grow. The company chose a suite of Azure services, including Azure AI Video Indexer, Azure Kubernetes Service (AKS), Translation, Azure OpenAI Service, Azure AI Language Studio, and Azure AI Services. 

As a product-led software as a service (SaaS) company, Lucy turned to Microsoft Azure to maximize the value of its investment with technology that does more. Lucy wanted to revolutionize Lucy to continue to meet the company’s mission “to amaze, delight, and empower people through knowledge.” 

The Lucy team could build their own copilot and generative AI applications, connect the data, call functions, enhance security, and improve workflows with language and image models. Lucy uprooted its infrastructure and moved to Azure during a three-month period in 2023 to help fulfill its mission. Lucy depends on the high availability, agility, and enterprise-level security of Azure.

“We make a promise that you can find information you didn’t even know existed in your company,” explains Steve Frederickson, Chief Product Officer at Lucy.AI. “Lucy is the fulfillment of that promise.” 

Lucy works across industries and departments making everyone more productive by eliminating the thousands of hours teams waste searching shared drives, scouring documents, distracting colleagues, and re-creating resources that already exist.

Enhancing Lucy’s competence

One challenge has been that video is incredibly hard to search. Now that Lucy has implemented Azure AI Video Indexer, she can draw from the source, and assess and prioritize video content alongside other formats such as PDFs and PowerPoint presentations. With Azure AI Video Indexer, Lucy can zero in on the exact moment in a video or webinar that answers the information seeker’s query. Since Lucy deployed Azure AI Video Indexer, it has been adopted by 100 percent of customers who’ve tried the solution. Plus, Azure AI Video Indexer answers get 55 percent more engagement compared with other first-party content in Lucy.  

The deployment capabilities and scalability of Kubernetes, specifically AKS, have proven critical to the successful evolution of Lucy. Frederickson and his small developer team are deploying as many as four new builds every day to four different environments—production, pre-production, testing and validation, and development. AKS “lets us manage what’s where and gives us the ability to roll back and branch into different categories,” says Frederickson. “And when we launch with a new enterprise customer, we need to index terabytes of data from thousands of SharePoint sites immediately. AKS is key to our being able to do that—and to dialing up our service levels on a launch announcement day.”

Lucy has a laser-like focus on realizing its corporate values of shared ownership, evolution, respond with urgency, and experimentation. For Lucy, shared ownership means that as many people as possible create a great product where, in addition to finding answers, Lucy is accessible, nurtures human relationships, and fosters connections with co-workers.

Lucy constantly evaluates whether its answer engine Lucy solves problems for individual users, evaluating success by considering enterprise employees with a consumer lens. It measures success using a Net Promoter Score (NPS). The spring release of Lucy scored a 15-point increase in the NPS. 

The future of Lucy

The Lucy team continues to explore where they can go by partnering with Azure. “My job is to create innovative apps with AI. The Azure toolset is top tier, and we are excited to explore where we can continue to evolve with a shared vision and roadmap,” notes Frederickson.

Lucy looks forward to the future and evolution of generative AI, with the goal of becoming an employee’s constant companion for knowledge that can empower them. As Frederickson says, “We believe in technology not to replace people, but to enable them to learn things they didn’t know existed, do their jobs better, make smarter strategic decisions, and be more informed.”

Find out more about lucy on Twitter, Facebook, and LinkedIn.