Claude & Claude Code training - put agentic AI to work
A live, hands-on half-day course delivered via Microsoft Teams by London-based practitioners. Learn to run agentic workflows with Claude and Claude Code - delegating real coding, research, reporting and automation work to AI you stay in control of.
- 3-hour live session on Microsoft Teams - next session Fri 21 August 2026, 9:00am to 12:00pm
- Morning or afternoon sessions, UK time - full recording included
- £150 + VAT per person - certificate included
- Small groups of up to 12, so your questions get answered
The four jobs AI agents are great at
Most people treat Claude like a smarter search box - ask a question, copy the answer, move on. That leaves the best 80% on the table. This course is built around the four kinds of work Claude does brilliantly when you let it act, not just answer:
Coding
Builds, changes and tests real software - from new features to hunting down bugs - working across your whole codebase, not one snippet at a time.
Workflow automation
Repeatable jobs become one-word commands anyone on the team can run - reporting, triage, document production, the lot.
Research
Reads the docs, the codebase or the live web, and comes back with a briefing you can actually use - sources and all.
Data detective work
Pulls data from your systems, cross-checks it and flags what disagrees - before it bites you in a renewal or a report.
Every one of the four is hands-on in the session - and three of them need no code at all.
Why choose this Claude training?
It's immediately practical
Every technique is demonstrated live, then you do it yourself with guidance. You leave with working set-ups, reusable Skills and prompts you'll use the same day - not a folder of notes to decode later.
You learn from daily practitioners
Your trainers are working engineers who use Claude and Claude Code in production every day - including on the very website you're reading. Even the training deck was produced by Claude Code. Real workflows, honest answers about what works and what doesn't.
Small group, your questions
With a maximum of 12 participants there's room for your specific questions. Bring a real task you'd like AI help with - the hands-on practice works best on your own problems.
Who it's for
Developers
Delegate whole coding tasks and ship faster with an AI pair that never gets tired.
Technical leads and managers
Understand what agentic AI really changes, and roll it out to your team safely.
Marketers, PMs and analysts
Use agentic workflows for the research, reporting and documents that eat your afternoons.
Anyone keen to learn
Tried Claude but suspect you're scratching the surface? No prior experience required.
Dates, pricing and booking
Every session runs live online for three focused hours - morning (9:00am to 12:00pm) or afternoon (2:00pm to 5:00pm) UK time, via Microsoft Teams.
Chat, Cowork or Claude Code?
Claude comes in more than one shape. We'll show you where Chat and Cowork fit - but the main focus of this course is Claude Code, because it's the environment that gives you the most control, and with it the best results.
Claude Chat & Cowork
- Day-to-day thinking, writing and analysis
- Working with your cloud documents on the move
- One-click connectors to Gmail, Notion and your CRM
- Pre-built Skills for common tasks
The accessible starting point - covered briefly, so you know when it's enough.
Claude Code (the CLI)
- Deeper work with code, files and folders
- Reading APIs and scripting real tasks
- Multi-step jobs you delegate end-to-end
- Automations that run on a schedule
The most powerful way to work with Claude - where you'll spend most of the session.
And don't let "command line" put you off: Claude Code needs no coding skills and is far friendlier than it looks. If you can write an email, you can drive it - and by the end of the session, you'll be comfortable doing exactly that.
What is Claude Code?
Claude Code is a command-line tool that brings Claude directly into your terminal for agentic work - meaning Claude can act autonomously on complete tasks, not just answer questions.
Instead of copying and pasting between a browser and your editor, you delegate whole tasks from your own machine: Claude Code reads your codebase or files, makes the changes, runs the checks and reports back. It turns Claude from a chatbot into a working partner - and not only for code.
Why it changes the game:
- Terminal integration - work without leaving your own environment
- Autonomous delegation - hand over complete tasks, not just get snippets back
- Context awareness - Claude understands your project structure, files and data
- Compound productivity - write, debug, research, report and document faster, every day
Real tasks Claude Code handles:
$ claude "Fix the failing checkout tests and explain what was wrong"
- Writing complete features from specifications
- Cross-checking data between two business systems
- Researching a topic and returning a cited briefing
- Debugging complex issues in your code
- Turning repeatable jobs into one-word commands
- Generating documentation and reports automatically
What the session covers
Three modules across three hours - the same material we deliver privately to businesses, including at CEO level. Every technique starts with a live demonstration, then you do it yourself with the trainer on hand.
- Prompting that works first time: role, task, context, format - and when it's worth the effort
- The advanced moves: few-shot examples and XML structure for clean, reliable results
- Keeping long conversations sharp: context-window habits that stop drift
- Chat, Cowork or Claude Code: a clear sense of which tool fits which job
- Skills: turn your best prompt into a shared team asset - we build one together, in five steps
- MCP connectors: let Claude work directly with Gmail, Notion, your CRM and the live web
- Install and first run: one line per platform, then a safe sandbox to work in
- The agentic loop - plan, act, verify - and how to delegate like a manager, not a micromanager
- The cockpit controls: plan mode, extended thinking, model switching and never losing a session
- CLAUDE.md: the standing brief that makes every session start from your team's context
- The centrepiece walkthrough: two systems disagree about the same customer - Claude Code reads both APIs, reconciles the data, flags every mismatch, and the whole job becomes a one-word command anyone can run
- Going live safely: MCP servers for your real tools, secrets that never touch the chat, and guardrails your whole team inherits from one file
- Live build: we aim Claude at a real workflow from your week
- The traps that catch new users - and the fixes for each
- A prompt library your team actually uses, and the rollout habits that make it stick
- Automations worth a look: the morning brief, research agents, email triage
- Your AI policy: why silence penalises your most conscientious people, and the one-page fix
- Audit agents: QA that never forgets - the exact pattern we run on this website
- Your 30, 60 and 90-day plan, so today turns into a habit
Meldrick's ten Secret Sauce tips
Threaded through the session are ten advanced tricks most people never find - the habits that separate casual users from power users. A taste of what's on the menu:
Talk, don't type - dictate rich briefs three times faster than typing
The five cockpit controls - plan mode, model switching and more, the top gear most users never find
A standards repo - onboard AI agents and new starters with one sentence
Secrets that never touch the chat - keys stay in the vault, always
Guardrails in a file - decide the rules once, and every teammate's agent is born knowing them
Your agent, anywhere - long jobs that carry on while your laptop's shut
The other four you'll have to come for - and every tip ships in the take-home Secret Sauce sheet.
Proven on real work
This isn't a course assembled from other people's tutorials. The material comes straight from training we deliver privately to businesses - including 1:1 sessions at CEO level - and from how we run Red Eagle Tech ourselves.
In one private delivery, the centrepiece ran live: Claude Code read the API documentation for two business systems - a CRM on one side, billing on the other - matched every customer between them and flagged the records where the two disagreed. The whole job was then saved as a one-word command the team can run any week. That same walkthrough is Module 2's centrepiece here.
And it compounds: even the training deck itself was produced with the assistance of Claude Code, using the same MCP servers you'll meet in the session.
Four audit agents. 95 checks. Every page.
Claude agents make excellent internal auditors - a use case most people never think of. Every article on this website is scored by a panel of Claude Code audit agents before it ships, checking the click, the content, the strategy and the house style against written checklists. Humans hold the gates; the agents never get tired. It's the exact pattern you'll learn in Module 3.
Certificate and certification
Your certificate of completion
Every attendee receives a Red Eagle Tech certificate of completion covering the session's three hours of structured, instructor-led learning on Claude, Claude Code and agentic workflows.
You can record the session as three hours of Continuing Professional Development (CPD), subject to your professional body's rules - keep the certificate and the course outline as evidence for your CPD log.
Eyeing Anthropic's certification?
Anthropic's credential is the Claude Certified Architect - Foundations (CCA-F): a proctored, closed-book exam of 60 scenario questions in 120 minutes, priced at $99 USD per attempt as of its March 2026 launch (check Anthropic's own pages for the current fee). It validates production-architecture knowledge - it's a badge that proves what you already know, rather than training that teaches you. We're independent practitioners, not an Anthropic partner, so here's the honest comparison:
| Comparison point | This course | Anthropic's CCA-F |
|---|---|---|
| What it is | Live, hands-on training | Proctored multiple-choice exam |
| You spend | 3 hours practising on real workflows | 120 minutes being tested, closed-book |
| Best for | Learning to apply Claude to real business problems | Engineers proving architecture knowledge |
| You leave with | Working set-ups + a completion certificate | A digital badge |
They pair well: this session covers the practical ground the exam assumes - Claude Code workflows, MCP, prompt engineering, context management - so it's a strong first step if you're working towards the badge.
Frequently asked questions
Training a team?
We run private sessions for whole teams, tailored to how you actually work - the public course is the same material we deliver privately to businesses.
- Delivered online, or in person in London and across the UK
- Examples from your industry, your tools and your codebase
- Flexible dates to suit your schedule
- Governance, AI policy and safe-usage guidance built in
- Follow-up sessions available
Let's build a plan for your team
Tell us your team size and what you want them to be able to do, and we'll put together a custom proposal.
Talk to us about team trainingNeed more than training?
If you'd like us to build AI into your business for you - from practical automations to full agentic workflows - our AI solutions team can help with that too. And if you're setting the ground rules first, start with our free AI policy template.
Explore our AI solutions servicesReady to put agentic AI to work?
Next session: Fri 21 August 2026, 9:00am to 12:00pm. Three hours now, hours back every week after.
Ready to discuss your project?
Book a free discovery call or send us a message. We'll respond within 2 business hours.
Discovery call
A friendly 15-minute video call with Kat to understand your needs. No preparation needed.
- Discuss your project
- Get honest advice
- No obligation
Kat Korson
Founder & Technical Director
Our team has 10+ years delivering software solutions for growing businesses across the UK.
Send us a message
Your information is secure. See our privacy policy.
Next: Fri 21 August 2026
Find us
