Back to case studies
    Knowledge Management

    Teams knowledge bot for service and field teams

    A RAG-powered Teams bot answers questions from SharePoint documentation inside existing workflows – with source citations, no new logins.

    50%

    funding possible

    < 60 Sek

    field response time

    0

    new logins required

    Problem

    Knowledge silos and back-and-forth communication slow down field operations. Service technicians must call colleagues or search document folders for every unknown problem – a time loss that directly impacts customer satisfaction.

    Solution

    Knowledge documents from SharePoint are structured in a RAG pipeline and made accessible via a Teams bot. Employees ask questions directly in the Teams chat and receive accurate answers with source citations within seconds – without leaving their familiar work environment.

    Tech stack

    Microsoft TeamsSharePointAzure OpenAIMicrosoft Entra ID

    Results

    • Faster answers in service operations without escalations to central teams
    • Fewer escalations and waiting times in field operations
    • Seamless integration into Microsoft Teams without new app installation
    • Funding support of up to 50% depending on program and region

    Running Costs

    The Teams knowledge bot runs inside the customer's own tenant. This means infrastructure, data, and AI usage remain entirely within the company's existing Microsoft and Azure environment.

    Direct technical operating costs stay manageable. Azure infrastructure typically amounts to around €30 per month. Usage-based costs for AI queries were approximately €10 per month during the pilot with 10 users.

    No additional hosting or platform fees are charged by Hoch-AI. Ongoing operations can be supplemented with a separate support package if needed, covering technical support, quality assurance, minor adjustments, solution monitoring, and regular check-ins.

    Further development after the pilot is handled separately. New features, additional data sources, more user groups, or integrations with existing systems are scoped together and then implemented. This keeps technical operating costs transparent while allowing larger expansion stages to be decided and budgeted deliberately.

    Curious how a project like this is structured from start to finish?

    Our process