With the rise of AI technology throughout the IT industry, Pericror has implemented a Natural Language Processing tool to mine key topics found within your data. Whether you decide to use our convenient, quick tool, or a internal data science resource is your choice- but in both cases, you should ensure to take action based on the results. Understanding the topics that are frequently found in your organization will help you:

  1. Identify Strategic Roadmap Decisions & Automation Investments
    • Once you understand the topics that make up the issues you face, you can understand high level project strategy. Are password reset and account lockout topics high on the list? Perhaps it’s time too invest in a SSO or password reset solution. Are users complaining about a number of privilege issues? Perhaps it’s time to invest in a self service solution to grant access, or provision needed software.
    • Having a discussion with your ServiceNow account contacts can help you identify additional services (like orchestration, password reset, etc.) that can help you solve the identified issues presented by the topics
  2. Identify knowledge gaps in your knowledge management system
    • An immediate action item is to ensure that you have self service accessible knowledge articles that cover the topics related to your users issues. Is VPN connectivity high on the list, but no articles on troubleshooting VPN access? Use the topic list to guide what articles are needed.
    • Maybe an article exists that should address the given topic, but it’s not being surfaced? Utilize the ‘Meta’ field found on the knowledge article and ensure to enter all the keywords from the associated topic.
  3. Identify Use Cases to utilize in a chatbot or virtual agent
    • Most virtual agents and chatbots today require manual definition of the conversations the bot should handle. From requiring training examples when doing intent/entity extraction via NLP, to actually defining the dialogs that a bot should be able to support, the manual configuration
    • Since the implementation is costly, it is important to prioritize building conversation support for your highest ranking topics

Finally, realize that different insights are available in different tables and contexts. A incident table’s short description field will indicate common issues, while it’s close notes field will indicate common resolutions. A change request tables short description will indicate common changes, while a chat tables message field will indicate commonly chatted topics. Make sure to run insights across many sources in your organization to identify larger trends.