Rackspace’s Advanced LLM-powered Chatbot

LeewayHertz developed an LLM-powered chatbot for Rackspace to optimize their internal support operations, accelerate response speed and elevate query resolution effectiveness. Leveraging NeMo Guardrails and other robust AI advancements, the chatbot assists users in performing specific tasks like creating support tickets and reducing the manual workload of users.

The Requirement

Rackspace, a leading provider of multi-cloud solutions, confronted a critical challenge with its service desk automation bot, the primary interface for internal team queries. The bot grappled with issues such as session drops and conversation delays, leading to suboptimal user experiences. Whenever the bot encountered queries exceeding its capabilities, it was thoughtfully either redirected for further assistance or dropped by the support team. Recognizing the need for improvement, Rackspace prompted a collaboration with LeewayHertz to develop a chatbot integrated with the capabilities of LLMs to improve query resolution efficiency, enhance support operations, and streamline the internal team helpdesk functions.

The Solution

LeewayHertz developed an LLM-powered chatbot thoroughly trained on an extensive ServiceNow knowledge base to optimize Rackspace’s internal helpdesk efficiency and support workflows. The integration of Retrieval Augmented Generation (RAG) techniques and Nemo guardrails empowers this smart chatbot to excel in validating responses and ensuring accuracy in user interactions. In scenarios where queries exceed its capabilities, the chatbot autonomously creates support tickets, making it a beneficial resource for Rackspace support operations. With its user-friendly interface integrated into Teams, the chatbot simplifies the user experience for employees. Augmented by prompt engineering techniques, the chatbot ensures the delivery of context-aligned and relevant responses, providing an optimized user experience.