GPT To Boost Knowledge Management Adoption

last updated:

GPT To Boost Knowledge Management Adoption

last updated:

At Deep Analysis, we keep up with the pace of knowledge management innovation, so you don’t have to. To the IT mainstream, KM has long been an esoteric corner where it’s difficult to pin down the actual business value – and that has held back enterprise adoption. That’s about to change with the rise of the LLM.

…with a little help from its friends

Daily we hear this question from our friends and clients: “Sure, ChatGPT is a crowd-pleasing demo; but what are the real business use cases?” So, it was perfect timing to be briefed by a company who already has customers using a similar product and has felt the hot breath of ChatGPT breathing down its neck.

We spoke last week with the folks from about the impact of ChatGPT on knowledge management for enterprises. The company has been using OpenAI GPT components for two years as part of its own solution. Even as we chatted, the product team were in the midst of an intense discussion over what the latest OpenAI API announcement could mean for their products.

First, let us familiarize you with Lucy, a software product one could think of as the “company answer machine”. To begin the process, Lucy crawls through enterprise files no matter where they reside: SharePoint, other repositories, Salesforce, SAP, file shares, etc.; ultimately any file source to which it is granted access. The software uses advanced NLP to tag and index the files while leaving them in place, and knowledge graph technology to understand context and relationships between words and phrases. Once Lucy connects to a data or file source, it becomes persistent so that any new content added to the file source will be indexed and ready for queries.

Employees can ask questions of Lucy using natural language queries, similar to the now-familiar ChatGPT prompts. From the indexed company knowledgebase, Lucy provides detailed and summarized answers in seconds, for questions ranging from “what is our corporate strategy for sustainability” to “how many female customers in Asia are buying our latest beverage.” Or anything else relevant to the company’s collective knowledge and experience. (For a deeper analysis of Lucy, please read our Vendor Vignette.)

While on the surface the end user experience may sound like ChatGPT, Lucy does far more – and also does it differently to the benefit of business.

The first and most important distinction is the source of the training data. In stark contrast to ChatGPT, Lucy’s large language model (LLM) is trained solely on trusted company information and the answers are inherently reliable. The company was quick to point out Lucy is not always 100% correct; but at least her answers are not based on the unverified and potentially ruinous training data that plagues ChatGPT and other internet-trained LLMs.

When the human thinks Lucy could do better, there’s a feedback option similar to the thumbs up/thumbs down in ChatGPT. Crucially, Lucy’s feedback mechanism can be limited to only verified subject matter experts (SMEs) whose suggestions can be trusted. Contrast that to ChatGPT, where any idiot can provide training input. Along with the answer, Lucy also presents the source documents from which the answer was derived and can even guide you to the page where the information resides.

We asked how their partnership with Microsoft will be affected now that the Redmond-based company has embraced GPT. The company believes Microsoft will move quickly to inject GPT into every product and add significant enterprise value to discover company knowledge. Today, SharePoint libraries are not easy to search with its current tools; that feature gap spawned a SharePoint search and knowledge management ecosystem of vendors who built their solutions on top.

When Microsoft finally gets SharePoint search right using GPT, the team believes this is a game-changer for those vendors and could even mean lights-out for those who fail to add value. We think Lucy’s product differentiation and innovation can keep them one step ahead of Redmond. One example is Lucy’s ability to perform a true enterprise search across all non-Microsoft repositories and return the results alongside information from the Microsoft siloes.

However, the company agreed with us that the speed of innovation for AI is far faster than anything we’ve ever seen in the software business, and no company can hope to survive without continuous innovation. Sam Altman, co-founder and CEO of OpenAI, recently tweeted that we’re on the brink of a new Moore’s law: artificial intelligence will double every 18 months. The team offered a pithier observation: AI software innovation is no longer measured in dog years; now it’s in hamster years. 

At Deep Analysis, we keep up with the pace of KM innovation, so you don’t have to. To the IT mainstream, knowledge management has long been an esoteric corner where it’s difficult to pin down the actual business value – and that has held back enterprise adoption. That’s about to change with the rise of the LLM and Microsoft’s integration of GPT into its VIVA platform. One thing’s for sure: no matter what happens, thanks to ChatGPT every business is now talking about knowledge management, whether they realize it or not. The rising tide could lift many boats, including that of



Leave a Comment

We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume that you are happy with it.

Work Intelligence Market Analysis 2024-2029