Claude Code Deep-Dive
Full day
Your codebase. Your problems. Hands-on session where your team learns to use Claude Code on real tasks — not toy examples.
Departs
Your developers are already using AI. The question is how.
84% use AI tools. Claude Code has 115K developers writing 195 million lines weekly. But a controlled study shows tasks take 19% longer without proper technique.
Stack Overflow 2025 · Anthropic Jul 2025 · METR 2025
Another prompting webinar — cancelled
Up to 100% funded through KFS
These workshops qualify for Krajowy Fundusz Szkoleniowy (KFS) government funding — up to 100% subsidy. I help with the paperwork.
What does a 33coders AI workshop look like?
- Scoping call
- 30 minutes to understand your stack, your team structure, and what problems you actually want to solve. I need to know your codebase, your deployment process, and where the bottlenecks are before I build anything.
- Custom preparation
- I build exercises around your actual codebase and toolchain. No generic examples. If your team works in Python with FastAPI, that is what we use. If it is TypeScript with Next.js, same. The exercises reflect your daily work.
- Workshop day
- Maximum 15 minutes of slides, then hands-on the rest of the day. Your tools, your problems, your code. Participants work on real tasks they will face next week, not toy problems they will never see again.
- Follow-up support
- All materials delivered the same day. Then 2 weeks of async support — your team can ask questions via email or Slack as they start applying what they learned on real work.
What makes these AI workshops different from a prompting course?
AI Tools Evaluation
Half day
Security audit of what your team is already using. Standardization plan. Clear recommendation on what to adopt, what to drop, what to watch.
Departs
Process Automation with AI
Full day
Workflow redesign, not chatbots. We identify repetitive processes in your engineering pipeline and build AI automation that saves measurable hours.
Departs
- An AI coding workshop is a hands-on, instructor-led training session where engineering teams learn to use AI development tools on their own codebase and real-world problems.
- Unlike generic prompting courses that teach ChatGPT basics, a proper AI workshop focuses on the specific tools, languages, and workflows that the team uses daily. Participants work through exercises built from their actual code, learning techniques for effective AI-assisted development that they can apply immediately. The goal is not theoretical knowledge but practical skill that measurably improves development speed and code quality from the first week after the workshop.
Who this is for
- DepartsEngineering teams of 8-15 people adopting AI coding tools like Claude Code, Copilot, or Cursor and wanting structured methodology instead of individual experimentation
- DepartsTech leads and engineering managers responsible for standardizing AI tooling across teams and needing a clear evaluation framework
- DepartsCompanies with 10-200 developers where productivity gains from AI are expected but not yet measured or realized
- DepartsTeams that have tried AI tools informally and got mixed results — some faster, some not — and need to understand what actually works
Who this is not for
- CancelledNon-technical teams wanting an introduction to ChatGPT — I work with developers, on developer problems, in developer tools
- CancelledCompanies looking for off-the-shelf e-learning content they can play on repeat — every workshop is custom-built for the specific team
- CancelledTeams that need AI strategy before AI skills — if you do not know what to build with AI yet, start with the leadership advisory
Before, during, and after
Before
Context brief — I learn your stack, your problems, your team structure. No generic slides.
During
Hands-on from minute 15. Your codebase, your tools, your workflows. Maximum 15 minutes of presentation.
After
All materials delivered. 2-week async support window for questions as your team starts applying what they learned.
The productivity truth
"10-30% faster" is a myth — unless you measure it right. A METR study found experienced developers actually took 19% longer with AI on real tasks. The difference is technique, not tools.
METR, 2025
Common questions
- How many people can attend a workshop?
- Optimal group size is 8-15 people. Smaller groups get more individual attention. Larger groups lose the hands-on element that makes these workshops valuable. For bigger teams, I run multiple sessions.
- Do we need specific tools installed beforehand?
- I send a detailed setup guide one week before the workshop. For Claude Code sessions, participants need an active subscription. For tools evaluation, bring whatever your team is currently using — the messier the landscape, the more useful the audit.
- Can the workshop be delivered remotely?
- Yes, all workshops work fully remote via screen sharing and collaborative coding. I have run over 20 remote sessions. The only requirement is that each participant has their own machine with the tools installed — no watching someone else code.
- How does the KFS funding process work?
- KFS (Krajowy Fundusz Szkoleniowy) can cover up to 100% of the workshop cost for Polish companies. I help with the full application process — from preparing the training program documentation to filling out the forms. The application needs to be submitted to your local labor office (Urząd Pracy) before the workshop date.
- What if our tech stack is unusual or proprietary?
- I adapt all exercises to your specific stack. During the scoping call, I learn your tools and build custom exercises. I have worked with teams using everything from mainframe COBOL to cutting-edge Rust. The AI patterns transfer — the examples just need to match your reality.