Getting Started with Lucy

OVERVIEW

Generation Lucy is an artificial intelligence (AI) powered knowledge management platform. Lucy exists to amaze, delight, and empower people through knowledge. She does this by driving data democratization, knowledge sharing, and greater access to historical brand data and 3rd party resources. Lucy’s AI reads, listens, and watches every document, audio, and video file from the clients’ connected SharePoint sites and can pinpoint specific answers based on natural language questions. Lucy unifies the data an organization owns or licenses through a single interface, so that she can serve as a GPS to insights and answer any question she has data to support. Lucy is a team-member who never leaves, never forgets, and becomes smarter every day through user feedback.  Synopsis is a Natural Language Generation

View full size  Figure 1 – Lucy’s Redesigned User Interface.

WHY IS LUCY IMPORTANT?

The need for knowledge management platforms, like Lucy, is born out of the challenge of data access in the face of the exponential growth of data available to marketers and research insight professionals. For most organizations, data is often siloed by brands, divisions, and geography, files are stored across the enterprise and it can originate from many sources, internal, research partners, or agencies. Further, when employees change roles and / or depart their historic knowledge is lost. These common challenges make re-use and discoverability of valuable data very difficult.

The goal for marketers today is to quickly find the data that matters and put it into action. With current tools though, the result, too often is a frustrating task of finding, collecting, and synthesizing data that can take days. It’s frustrating to know the right data is out there, somewhere. This undiscovered knowledge costs time, wastes money, and affects employee performance.

The reason to adopt a knowledge management platform is to find the right answer, right when you need it. Lucy is built for the Fortune 1000. Insights team-members and marketeers can find answers within the huge amount of unstructured data at any organization (e.g. PowerPoints, PDFs, Word Docs, audio, and video) across brands and geographies. Scenarios include saving time while trying to find that chart, research, study, analysis, or other needed data to support a variety of work, and Lucy can save endless hours reducing the duplication of work that has already been done. Lucy is here to solve these problems and support the insights community.

Today many organizations' existing insights are underutilized:

• On average the researchers spend 8hrs per week searching for data.
• 73% of enterprise data is going unused across similar organizations.
• Furthermore, after 90 days many knowledge assets will never be used again (up to 90% of this data).

WHAT BUSINESS QUESTIONS CAN LUCY ANSWER?

Via natural language questions Lucy can return answers to many business questions. She is most effective at finding specific answer(s), versus a list of possibilities, bringing users to the exact spot within the source where the answer lives—for example a chart within a presentation, the precise second in a video where the answer is referenced, or the specific paragraph in a multi-page document. Lucy answers these many questions by leveraging, and learning from, all the content she’s been connected to. Lucy’s connections are automated, thus anytime new information is added or updated she learns that too. She can also be connected to third-party data sources such as Mintel, Euromonitor, Kantar and more!

Other common questions that Lucy can help answer:

• What does the insights community know? (i.e., Reduce dependence on subject matter experts).
• What research exists in different markets? (i.e., Share insights across the organization).
• Can we save duplicative briefs? (i.e., Eliminate redundant research).
• What research have we done before? (i.e., Drive employee efficiency).
• Where did these assets go? (i.e., Retain institutional knowledge).

HOW DOES LUCY WORK?

If your search habits were formed by systems using labels or tags, you’ll be surprised at how much freedom you have when asking Lucy questions! Natural language processing allows her to understand queries like a human. Simply put, when you are looking for something, just type a question as if you were asking a co-worker. To get started, ask her a natural language question to support a business need or a daily task:

View full size  Figure 2 – Keyword are for cavemen! For best results, use natural language search questions.

Lucy will quickly present you with the 10 best answers she feels relevant based on the information she has available. Answers are prioritized by relevance and shaped by user feedback. If you did find what you’re looking for make sure to let her know so she can get smarter. All feedback, whether positive or negative, will help her continually learn and will benefit the entire organization. Lucy learns through user interactions, with her primary feedback mechanism through the 👍 👎 available within every answer.

If you don’t find what you’re looking for right away, try a few of the tips below:

• Try asking the question in a different way – using different keywords or question framing can unlock other insights.
• Be as specific as possible – additional context helps her identify what is relevant.
• Leverage filters – focus results by filtering by Source, File Type, Timeframe, or Content which limits the content being searched. Filters are powerful tools when looking for content on a brand, from a SharePoint, or excluding dated information. If you choose to use keywords to search, it’s recommended to use Lucy’s Content filter to help refine the broad results returned.

If you still don’t find what you’re looking for let Lucy know. After every question she will ask, “Did you find what you’re looking for?” Providing this feedback helps to improve her understanding of your domain and ensure no content is missing.

Like a new team-member, Lucy is only as good as the information she has access to and the training she receives. If you are unsuccessful in finding content to support your topic of interest, brand, or information from a particular region, it’s possible Lucy hasn’t been provided access to the content to support your question. If this is the case, follow the steps in the Key Facts section to connect Lucy to the data important to you and your team.

KEY FACTS

• Lucy is a Learning System: Today is Lucy’s worst day. She will never know less than she does today. Tomorrow, she will be smarter and the day after smarter still. She gets smarter every day on the job based on the usage of the system and the feedback provided. Together, we are raising an AI that will become amazingly smart over a relatively modest period of time—so long as we onboard our new team member effectively!
• Use Natural Language Questions: For Lucy, it all began with natural language search. The idea is you talk to Lucy and ask questions as you would between two co-workers. This is ideal when you believe that the data to answer the question exists.
• Lucy Delivers Answers Not Search Results: While other systems provide lists of documents or summarization of data sets, Lucy provides a specific answer(s) and brings users to the exact spot within the source where the answer lives.
• Find Documents Differently: Lucy focuses on finding the precise answer first but always has the list of documents available via document view. After asking a question, simply select the “document” icon to the right of the filters to see the list of documents she found the answers presented within. Additionally, a specific document can be found through document view by entering “filename: XYZ.pdf” in chat if part of the file name is known.
• Lucy Understands Your Domain: Through Lucy’s taxonomy expansion capabilities, she is programmed to understand associations specific to your language and your brands, immediately improving accuracy.
• Lucy Accesses Organizations' Knowledge Where It Lives: Lucy operates with the data in its current environment versus having to move the data into a knowledge system. Other solutions recommend developing a “Librarian” role to manage the flow of data into their knowledge portals.
• Save and Share Your Favorite Answers: Need a place to store helpful answers when working on a project? You can save any answer to your bookmarks by clicking the bookmark icon in the upper right-hand corner of an answer. To share an answer with anyone in your company, click the share icon in the upper right-hand corner of the answer.