Seeing What Text Analytics and Gen AI Can Do for You

Text analytics can create a platform that can realize the full value of Gen AI in the enterprise where it is struggling to create real business value. However, building this platform can be difficult and justifying the expense and labor can be a complex task.

The KAPS Group, in conjunction with our partner text analytics vendors, have developed a simple but powerful approach that clearly demonstrates the value of the pairing of text analytics and Gen AI, and at the same time, takes a first step toward building a text analytics-Gen AI platform.

The heart of the technique is a mini-POV that in a week builds a demo using your content and our expertise and software.  The process is simple and requires only a small effort on a client’s time and resources. 

Seeing is Believing

The primary obstacle slowing the adoption of both text analytics and Gen AI in the enterprise is a lack of understanding of what it is, what it can do, and how best to implement it.

A Mini-POV provides answers and guidance for all of these concerns. Specifically, a Mini-POV:

• Creates something that everyone can see in action with demonstrable results 
• Creates a real application with real content
• Can appeal to all audiences from librarians to KM practitioners to IT to executives
• Starts the building of a foundation for multiple text analytics and Gen AI-powered applications

A Mini-POV is something that can benefit any organization regardless of their current text analytics or Gen AI understanding or implementation. It can:

• Introduce the power of the pairing of text analytics and Gen AI to organizations that are new to the concepts, techniques and applications
• Demonstrate how to build an enterprise LLM that reduces hallucinations   
• Demonstrate the value of moving beyond the generic and too often wrong Gen AI answers combining text analytics and RAG search-based applications
• Demonstrate how to take an existing text analytics capability to a new level by building flexible rules that utilize document structural elements
• Gain an understanding of the scale of effort for a full text analytics initiative

Mini-POV Process

A basic Mini-POV takes one week of concerted effort that can be spread out over a number of weeks to accommodate schedules.  It doesn’t require the installation of any software.

• Step One is selecting 10 categories from a current taxonomy or creating 10 new, high-level topics

• Step two is to work with the client to select 20 documents per category for a total of 200

• Step three is for the KAPS Group to build a simple content model that captures any document structure (and all so-called unstructured content has some structure)

• Step four is for the KAPS Group to build categorization or auto-tagging rules that achieve over 90% accuracy

• Step five is to apply the rules to the output of selected Gen AI prompts to correct hallucinations

• Step six is to demo the results to selected audiences within the organization

The demo showcases the greatly improved accuracy over traditional search results as well as providing a reality check on Gen AI.

Options: The Mini-POV can be done as a stand-alone project or can be combined with the KAPS Group strategy workshop: Understanding Text Analytics and Your Organization, which provides a strategic look at how text analytics can be utilized throughout an organization and integrated with Gen AI and existing information and knowledge management initiatives.

Pricing: This engagement is a fixed price offering of $10, 000 although it can be expanded if the client would prefer a more in-depth engagement of 2-3 weeks.  The expanded engagement allows for more content and additional use cases.

If this sounds like something that you would like to learn more about, please contact Tom Reamy – tomr@kapsgroup.com