Faced with underperforming chatbots that were threatening to compromise employee experience and productivity, a large tech manufacturer turned to Pryon to evolve its IT support strategy. Now, Pryon transforms more than 500 documents into answers to more than 20,000 questions per month with 90 percent accuracy — three times the accuracy of its previous chatbot — and estimates $4.4 million in savings from support ticket reduction. Pryon’s no-code AI capabilities make it easy to automate exceptional employee support while allowing the workforce to focus on strategies and innovations that keep the company competitive.
IT help desks can struggle to keep pace with employee requests, especially as remote working has increased the volume and complexity of hardware, software, and security demands. The company had launched a chatbot to scale and automate as much of their IT support as possible but it wasn’t reaching the desired levels of automation and accuracy to justify the effort and expense.
Pryon was brought in as the AI-powered link between the chatbot and the company’s internal knowledge bases, which reside in ServiceNow and Confluence. The current ServiceNow search functionality is limited to keyword search, which means users need to know the exact word that the system recognizes in order to produce an answer. Pryon, on the other hand, uses natural language processing (NLP) engines to understand variations of words, questions and phrases in order to intelligently deliver the right answer. For this application, Pryon provides all of the chatbot’s answers by reading and transforming the knowledge base content for employees — and the API integration provided a way to rapidly accelerate time-to-value.
Pryon’s ability to display the exact location of the answer within the company’s source document saves users time (compared to traditional search results, which deliver a list of links that users must painstakingly click through) and also increases the user’s confidence in the quality of the answer. Users can then click to validate that the answer is accurate, which improves the speed and quality of future user sessions related to that same question.
Subsidiary support during acquisition
Having seen the rapid deployment and time-to-value of Pryon IT support application, leaders at this manufacturing company consulted with the Pryon team about its recent acquisition of a computer hardware distributor. The period after an acquisition or merger is a critical make-or-break time for customer experience — if customers have a negative interaction, they may leave and never come back.
The target company had come with a web-based support chatbot that the parent company team wanted to integrate into their new Pryon-powered chatbot to streamline the digital experience as the target’s website was being consolidated and deprecated.
The flexibility and full-stack nature of the Pryon platform, along with the deep level of commitment and experience of the Pryon team, brought this project into completion well within the aggressive 30-day deadline.
Pryon’s no-code AI capabilities make it easy to automate exceptional employee support for this industry-leading manufacturer, while allowing the workforce to focus on strategies and innovations that keep the company competitive.
Next up: Pryon to transform customer contact center performance
Despite the fact that this manufacturing powerhouse has invested in multiple technologies to manage customer support, the automated answers being delivered to customers via its current chatbot are often inaccurate or incomplete — even though the company has a rich set of content and web-based documentation with all the necessary information. Missing is a highly intelligent and automated way of transforming that content into accurate answers at scale.
In a recent pilot for the customer support organization, Pryon’s expert solution engineers deployed the Pryon Knowledge Operating System and were able to quickly ingest the relevant internal documentation from its existing locations, create knowledge collections organized by product and service categories, and define user access levels.
The first wave of question-and-answer testing showed a 3x increase in accuracy compared to tests without Pryon. This means that three times more answers were accurate, and pointed users to the precise point in the documentation where the answer was found.
The theme: Rapid, repeatable time-to-value
At a time when traditional AI projects are taking months, even years, to deploy — and even longer to prove value — Pryon is setting new standards for time-to-value with its low-overhead, high-return approach.
No need for a corps of data scientists. No need to move or reconfigure your content. No manual chatbot programming. Just point-and-click connectivity with immediate payback, and a rapid, repeatable way to show momentum in your digital transformation efforts.
