AI HackathonsConsultingAbout
    Anthropic Claude logoInstructor-led team training

    Claude AI Training Course

    A hands-on, one-day Claude course for UK teams who want to get useful work out of Claude — built around your team's own tasks.

    In short

    • For UK businesses training 10 or more people
    • Around 16 people per workshop, with multiple sessions for larger groups
    • Delivered on-site anywhere in the UK, or live online
    • A scoping call confirms tools, licences, tasks and format
    • Price depends on group size and format

    About this course

    A serious, hands-on Claude AI training course for UK teams and professionals who want to get useful work out of Anthropic Claude. One day, with workshops of around 16 people and multiple sessions for larger teams, built around your team's actual tasks — covering prompting, Projects, Artifacts, long-document workflows, Skills, connectors and safe use.

    Course details

    Duration

    The standard course is one full day, delivered as roughly six hours of instruction and hands-on practice. It can also be split into two half-days for teams who cannot lose a whole working day, or condensed into a half-day introduction.

    Group size

    Workshops usually run with around 16 people so everyone gets time with a trainer during the exercises. Larger groups are run as multiple sessions.

    Delivery formats

    In-person at your offices anywhere in the UK, at a venue in Leeds, or live online for distributed teams.

    Lead time

    We usually need two to three weeks between the first conversation and delivery. That covers a short scoping call, gathering example tasks from your team, and tailoring the exercises. Faster turnarounds are sometimes possible — ask.

    See it in action

    What a Claude session actually looks like

    Choose a tool and explore a sample task, from brief to draft to human review.

    Claude logo

    A consultancy turning discovery-call notes into a client workshop proposal.

    Sample brief

    Brief

    Prompt — workshop proposal for Harlow & Finch

    Written by the account lead

    You are helping me draft a workshop proposal.

    Reader: Head of Operations. Decision: approve a pilot.

    Discovery call notes: 40-person ops team, monthly supplier reports take two days.

    Sections: context, proposed sessions, what changes, next step.

    Rules: no invented figures, flag every assumption, under 700 words.

    Explore the curriculum

    Pick a topic, see how it is taught

    Four of the topics covered in Claude training: pick one to see a sample exercise and what attendees take away.

    Working with long documents

    Sample exercise

    Attendees bring one long, non-sensitive document — a policy, a scope or a report — and work through summarising it for a specific reader and decision.

    What attendees leave with

    A reusable summarising prompt, plus a short note of what had to be checked by hand.

    What you'll learn

    • How Claude differs from ChatGPT and Gemini, and where it genuinely shines
    • Writing strong prompts and system prompts for Claude
    • Using Claude Projects, Artifacts and long-context windows effectively
    • Working with long documents, research and structured analysis
    • Turning repeated tasks into Claude Skills and reusable workflows
    • Understanding connectors and MCP, and what to connect first
    • Using Claude safely with company information and sensible guardrails

    How we deliver this training

    Instructor-led workshop

    A practical, hands-on session led by a Day Seven trainer. Run in-house for your team, or join an open workshop in Leeds. UK-based, jargon-free, and built around the work your team does.

    Get a training proposal

    Example exercises we can scope for your team

    We agree the exercises with you before the session, so they match the work your team does.

    • A document-analysis workflow: taking a long report, contract or tender and producing an accurate structured summary or comparison table, checked against the source
    • A Project set up for a live piece of work, with the right background context and custom instructions
    • One repeatable team task packaged as a Skill or saved brief, so the whole team runs it the same way
    • An Artifact your team can reuse — a template, table or small internal tool
    • A short governed-use note: what can be pasted in, what needs verifying, and who checks it

    What we need from you beforehand

    • A short scoping call to agree the tools your team uses, the tasks that matter and the format that fits your week.
    • Confirmation that attendees can sign in to the tool on their team's current plan or licences before the day.
    • Two or three real, non-sensitive examples per team — a typical document, report or email thread — so exercises use your own work.
    • A named internal contact for room, kit and joining details, plus anything your IT or security team needs to approve first.

    How to tell whether it worked

    • Agree at scoping which two or three tasks the training should make faster or easier, so there is something concrete to look at afterwards.
    • Compare the time taken and the quality of the output on one selected task before and after the session, so there is a performance measure alongside how people feel.
    • Capture a simple confidence rating from attendees before and after the session on those same tasks.
    • Ask managers a month later whether the agreed tasks are being done with the tool, and how often.
    • Keep the team's shared prompt library as a visible sign of use — an empty library a month on is a useful signal in itself.

    Our guide on measuring AI training ROI sets this out in more detail.

    Who it's for

    Knowledge workers, analysts and writers who want a sharper AI partner
    Product, design and engineering teams using Claude for thinking and drafting
    Bid, legal, policy and research teams working across long documents
    Founders and consultants using Claude for research and client work
    L&D and AI leads rolling out Claude alongside other AI tools
    Non-technical staff who have tried Claude once and stopped

    The curriculum

    Module 1

    Prompting and briefing Claude

    • How Claude interprets instructions, and why structure beats clever wording
    • Writing system prompts and reusable briefs for recurring tasks
    • Giving context, examples and constraints so output needs less editing
    • Iterating on a poor answer instead of starting again
    Module 2

    Projects: giving Claude persistent context

    • Setting up a Project for a team, client or ongoing piece of work
    • What belongs in project knowledge and what does not
    • Custom instructions at project level for consistent tone and format
    • Keeping projects tidy so answers stay accurate over time
    Module 3

    Artifacts and building small tools

    • Using Artifacts for documents, tables, diagrams and simple apps
    • Iterating on an Artifact rather than regenerating from scratch
    • When an Artifact is the right output and when plain chat is faster
    • Sharing and reusing what you build with colleagues
    Module 4

    Long-document and research workflows

    • Working across long reports, contracts, tenders and research packs
    • Structured extraction: turning a long document into a usable summary or table
    • Comparing several documents and spotting inconsistencies between them
    • Checking Claude's answer against the source before you rely on it
    Module 5

    Claude Skills

    • What Skills are and how they package a repeatable way of working
    • Choosing tasks in your team that are worth turning into a Skill
    • Writing clear instructions and supporting files for a Skill
    • Maintaining Skills so they do not quietly go stale
    Module 6

    Connectors and MCP

    • What the Model Context Protocol is, in plain terms
    • How connectors let Claude read from and act in your existing systems
    • Deciding what to connect first, and what to keep out of reach
    • Questions to put to IT and security before connecting anything
    Module 7

    Safe and governed use

    • What is and is not safe to put into Claude, with clear examples
    • How consumer, Team and Enterprise plans differ on data handling
    • Checking outputs: hallucination, confident errors and where the risk sits
    • Drafting a short internal usage policy people will actually follow
    Module 8

    Embedding it after the day

    • Choosing two or three real tasks per person to apply it to immediately
    • Building a shared prompt and Skills library for the team
    • Identifying internal champions and how to support them
    • A simple way to review what is working a month later

    What your team can do by the end

    • Write a brief for Claude that produces usable output on the first or second attempt, rather than after ten rounds of nudging
    • Set up a Project with the right background context and custom instructions for an ongoing piece of work
    • Take a long document and produce an accurate structured summary, and check that summary against the source
    • Build and edit an Artifact — a document, table or small tool — and share it with colleagues
    • Turn a recurring team task into a reusable Skill or saved prompt rather than rewriting it each time
    • Explain in plain language what a connector does and what questions to ask before enabling one
    • Recognise when an answer needs verifying, and know what information should never be pasted in
    • Explain to a colleague where Claude fits alongside the other AI tools your organisation uses

    Who delivers the training

    Training is delivered by Qasim Ahmed, founder of Day Seven. Qasim has 15 years leading digital learning teams across Aviva, KPMG and William Hill, and designs sessions around what actually works with real teams rather than what looks good in a demo.

    That background shapes the format. Sessions are hands-on, jargon-free, and built around tasks your team brought with them. We would rather your people leave with three things they will genuinely use on Monday than a tour of every feature.

    Day Seven is UK-based and delivers training in person across the UK and live online.

    Still deciding between Claude and ChatGPT for your team? We have written an even-handed comparison covering strengths, weaknesses, governance controls and cost at team scale — including where ChatGPT is the better choice.

    Read our Claude vs ChatGPT comparison

    Frequently asked questions

    Both. We start with the fundamentals of working with Claude, then move into Projects, Artifacts and longer workflows. Advanced users get more time on system prompts, Skills and chained tasks. If a cohort is clearly split, we tell you before the day and adjust the exercises.

    Around two thirds of the material is shared: how to brief a model, how to judge output, what is safe to paste in. The Claude-specific third focuses on depth over breadth — Projects, Artifacts, working across very long documents, Skills and connectors. ChatGPT training spends more time on model choice, custom GPTs and the wider feature set. If you use both tools, we teach the shared fundamentals once and add a short tool-specific session for each.

    Yes. The course is written for business users. If someone can use email and a word processor, they can follow the day. Nothing requires coding, and technical concepts such as MCP are explained in plain language with no assumption of a technical background.

    The full course needs a paid plan, because Projects and higher usage limits are used in the exercises. We check your team's current plan and licences on the scoping call and adjust the day to match — if you are still evaluating Claude, we can run a reduced version and revisit the paid features later.

    Yes, and that is how most of it is delivered. We come to your offices anywhere in the UK, or run it live online for distributed teams. In-house delivery also means the exercises can use your real documents and tasks, which makes a substantial difference to what people retain.

    Anthropic's free material is good and worth doing — it is thorough on how the product works. The difference is context and accountability. We build the day around your team's actual tasks, answer questions about your specific documents and systems, cover your governance position, and there is someone in the room when an exercise does not work. Free self-study is excellent for motivated individuals; instructor-led training is what tends to shift a whole team.

    Yes. There is a module on Skills — how to package a repeatable way of working — and a module on connectors and MCP covering what they do, what to connect first, and the questions to put to IT and security before enabling anything.

    Yes. We run a 'Mastering Claude' workshop in Leeds, and a live online Claude masterclass. You can also book in-house training.

    Yes. Day Seven is UK-based, in Leeds, and delivers Claude AI training in person across the UK and live online for remote teams.

    Yes, as a full module rather than a footnote. We cover what is safe to put into Claude, how consumer, Team and Enterprise plans differ on data handling, how to spot confident errors, and how to write a short internal usage policy people will actually follow.

    The final module is about making it stick: each person leaves with two or three real tasks to apply it to, a shared prompt and Skills library for the team, named internal champions, and a simple way to review what is working a month later.

    Explore other AI tool training

    Before you buy

    If you are still building the business case, our guides cover what AI training costs in the UK, how to choose an AI training provider and how to measure the return on AI training.

    Ready to build real Claude skills in your team?

    Tell us who needs training and what they work on, and we will scope a programme for your organisation.