deskdude.dk services
available Book a scoping call
Themes

    AI pilots and knowledge assistants for small businesses.

    I help teams buried in documents, repeated admin, or manual handoffs get a first AI system into use: an internal assistant, an automated workflow, or a knowledge assistant that answers from your own material with sources.

    Copenhagen independent builder-led delivery

    Choose the smallest AI build that can prove the value.

    Each package ends in a deliverable, so you know what you are buying and which decision comes after it.

    Not sure which package?
    scoping

    AI scoping and process scan

    Find the first realistic AI opportunity before spending money on a larger build.

    Fits
    Founders and leaders with many AI ideas, but no clear starting point.
    Timeline
    Typically 3-5 workdays.
    Deliverable
    Process map, 3 prioritized opportunities, risks, and recommended starting scope.
    knowledge assistant

    Knowledge assistant with sources

    RAG in plain terms: an assistant that searches your own documents and answers with source references.

    Fits
    Teams with manuals, guides, PDFs, web content, or internal knowledge spread across places.
    Timeline
    Typically 2-3 weeks.
    Deliverable
    Defined knowledge base, chat flow with citations, QA questions, and review.

    Smaller formats and ongoing help.

    Starting prices include a clear scope, setup, one review, and a short handover. Operations, API usage, extra integrations, and larger data volumes are scoped separately.

    AI workshopWhen the team needs to learn one practical AI work process.from DKK 9,500
    Ongoing AI opsWhen the first solution needs maintenance, testing, and improvement.from DKK 12,500/mo.
    Consulting hoursWhen the work is advisory, review, or technical sparring.DKK 1,250/hour

    Small cases with problem, solution, and result.

    Client cases can only be shown once approved for publishing. Until then, these internal builds show the same method: understand the problem, build small, test in practice.

    Restaurant Training

    Problem Pre-service training is uneven and hard to follow up on.

    Solution Mobile briefings, readiness, and onboarding.

    Result Active product with a narrow operating flow.

    Ask DeskDude AI

    Problem Answers require manual search across scattered content.

    Solution AI widget connected to site and materials.

    Result Live proof before client packaging.

    AI material in operation

    Problem AI knowledge stays too abstract.

    Solution Guides and test questions for real tasks.

    Result Reusable material for workshops and onboarding.

    How an AI project starts

    The process is designed to surface value quickly without hiding risk.

    01

    Map the task

    We identify the real work, data sources, failure risk, and smallest version that can be tested.

    You get: process map, scope, and decision on the next step.
    02

    Build the first solution

    The version is kept tight so it can be used internally and judged on real tasks.

    You get: working trial, test data, and known limits.
    03

    Test and move forward

    The system is reviewed with test questions, known limits, documentation, and next actions.

    You get: QA notes, handover, and recommended operations from here.

    Add-ons around the AI project.

    Use these as add-ons to an AI track, not as the main offer.

    Should we find out whether you need AI, automation, or a knowledge assistant?

    Book a scoping call, and I will identify the smallest useful solution.

    Book a scoping call