Overview
      Back to home
      1. Knowledge Base
      2. CrushBank User Guide
      3. Overview

      CrushBank Neuro End-User Guide

      Learn how Neuro can help your business find answers faster

      What is CrushBank Neuro?

      CrushBank Neuro is a powerful AI-driven tool that uses advanced search and retrieval techniques to assist IT professionals by answering questions with high precision and relevance. It leverages your organization’s data to provide accurate, contextual answers that are directly applicable to your work.  This guide will walk you through the core aspects of using CrushBank Neuro effectively, helping you understand how it works, its key features, and best practices for asking questions. 

      How does CrushBank Neuro Work?

      • CrushBank Neuro is built on a robust AI platform designed to retrieve information from your company's specific data sources. Using Conversational Search and Retrieval-Augmented Generation (RAG), Neuro processes queries to deliver precise answers drawn from your IT environment. Unlike generic AI models, Neuro is specialized in IT support and uses your own data for enhanced relevance. 
      • CrushBank Neuro allows you to chat with their organization’s documentation and extract answers to a variety of questions. Instead of having to query documentation for relevant results and then dig through each document for the answer, Neuro extracts the exact answer for you.   
      • All of the documentation within your organization’s data lake is broken down into chunks and indexed in a vector db. When you ask a question, Neuro searches across the vector database and extracts the most relevant chunks pertaining to the question. It then embeds those chunks along with the question within a finely tuned prompt and passes that to a large language model for processing which in turn generates the detailed answer. 

      What Are The Data Sources For CrushBank Neuro?

      All of your data sources from CrushBank are indexed into Neuro. (You can check what has been ingested in the admin portion of CrushBank)
      • PSA: ConnectWise, AutoTask, Halo, eAutomate, ServiceNow
      • Documentation: Sharepoint, IT Glue, Confluence
      • URL Uploads: Directly into CrushBank Pod or WebApp
      • API Ingestion: Directly ingest non-integrated data sources into CrushBank

      How Frequently Does Data Update Into CrushBank Neuro?

      Your CrushBank IT data is indexed into Neuro twice daily.

      Data ingested into CrushBank IT via integration generally updates hourly.

      Most ITGlue data updates hourly, but your Documents are ingested into CrushBank IT twice per week (due to limitations in ITG's API).

      How Do I Use CrushBank Neuro?

      Consider Neuro your librarian for your data ... it knows where all of it is, and can retrieve it for you!  

      • Microsoft Teams Chat: 
        • Neuro is accessible via Teams. Your organization must have a CrushBank account in order to use Neuro, and each Neuro user must have a CrushBank account. Once your organization installs the Teams app, you simply access the app within Teams or and type in your questions.
      • Ask a question
        • Ask it how to set up a new PC for this client, or how to set up that software for that client, or the IP address for the firewall at this location.  Anything you have documented and ingested into CrushBank can be returned in a conversational format.  
      • Ask a follow-on question
        • Neuro remembers your conversation from one question to the next.  You can continue the conversation and ask about setting up a new PC for a different client, for example, or you can "Start New Topic" to start over.
        • i.e.: "How do I set up Outlook on a PC?" "What about on a Mac?"
      • "Start New Topic"
        • If you'd like to start a new conversation and clear the memory of past questions, just click "Start New Topic".
      • "Query Tickets" button (coming 10/7/24)
        • If your documentation does not contain the answer you're looking for, you can click on the "Query Tickets" button to find that nugget of information buried in a ticket.

      What Is The Difference Between The CrushBank Pod And CrushBank Neuro?

      The CrushBank pod brings back search results from your data lake

      What Is An Aggregated Answer?

      An aggregated answer is when Neuro combines information from multiple sources to create a single, comprehensive response. Instead of pulling from just one document or ticket, it synthesizes data from several locations, giving a fuller picture and better context for solving a problem.

      Who Should Use CrushBank Neuro, And How?

      • Technical Staff
        • Configuration Information
        • How-To Documentation
      • Sales & Marketing
        • Marketing Collateral
        • Pricing Sheets
      • Account Management
        • Contract Information
        • Project Proposals

        What Are Best Practices For Asking Questions Of CrushBank Neuro?

        The better question you ask, the better answer you’ll get!

        Neuro's current use case is around retrieval of information.  Think of it as a librarian ... it will go out and find the answer to a specific question.  What's the IP address of this firewall?  How do I setup this software for this client?

        What Features Are Coming For CrushBank Neuro?

        As of September '24, we are working on Tickets and Documentation Request submission (expected 10/7/24), Neuro for Slack (expected 11/24), complex/aggregated answers (expected end-of-year), internet searching and knowledgebase building (expected early 2025).

        What Are Some Technologies Similar To CrushBank Neuro?

        CrushBank Neuro can be compared to generic AI systems such as ChatGPT or Microsoft CoPilot, but its IT Support-specific design sets it apart. Instead of the generic solutions which are not tailored to your internal systems, Neuro uses your own data to provide contextually relevant answers.