AI workshops for developers
From one-off prompts to a way of working you can rely on.
Your team already uses AI, but the results are uneven and the code still needs a lot of fixing. With the team in one room, we practise project context, planning, implementation and checking the results on a real project. The goal isn't more generated code. It's knowing how to use AI deliberately, and when not to use it at all.
- For
- Experienced developer teams
- Format
- In person, on a real project your team owns
- Length & group
- Agreed after a short call
First step
A short email about your team, your stack and what's frustrating.
bartek@bartekwitczak.com
Why AI results often disappoint
The problem is rarely the model. It's the work around it.
Too little context, or too much irrelevant context
→ The context the task needs, nothing more
Going straight to "write the code"
→ Explore the problem and options, then plan
One-off prompts that work once
→ A repeatable workflow the whole team uses
Domain knowledge only in people's heads
→ Domain knowledge written down for the model
No clear rules about what to build and how
→ Project rules the model can follow
Output checked by eye, or not at all
→ Verification decided upfront, automated where it makes sense
Lessons from each task are lost
→ What worked is captured for the next task
Weak tooling, sometimes no clear process
→ A setup that supports the workflow
The workflow we practise
One loop, repeated for each piece of work
Explore
Understand the problem, the code and the domain. Look at options.
Plan & specify
Agree what should change and how you'll know it works.
Implement
The agent works within clear rules and context, one step at a time.
Verify
Tests, automated checks and human review where it matters.
Capture
Save rules, context and mistakes to avoid back into the project.
↺ The next task starts with everything the project learned from the previous one. Not every task needs every step at full depth. Part of the skill is judging how much structure a change needs.
How the workshop works
No off-the-shelf package. Four steps.
Short email and call
Your team, stack, tools and what isn't working. We check if this workshop fits.
Preparation
We pick the project and agree tools, accounts, access and rules for code and data.
In-person sessions
I come to you. We adapt the rules and workflow to your project as we go.
After the workshop
If follow-up makes sense, we agree its form and scope in advance.
Why in person
It works much better than remote
- I see what’s really happeningI look at the screen, see where the agent went wrong and help right away.
- People learn from each otherComparing approaches at the next desk is where much of the learning happens.
- The team is fully focusedNo muted cameras, no Slack in another window.
- A shared way of workingRules and workflow are easier to settle with everyone in one room.
You leave with
Skills
- A method practised on your own code
- The ability to judge AI output and step in
- A shared vocabulary across the team
You leave with
Starting artifacts
- A first project setup for AI work: instructions, rules, context
- A written workflow your team can keep using
- Notes on what worked on your project
Boundaries
What this is not
- Not a prompt-tricks course. The bigger gains come from context, planning and verification.
- Not a change of your company's whole delivery process. That's AI in software delivery.
- Not a promise of "X% faster". Better, more dependable work is the outcome.
- Not a legal, compliance or security review.
- No guarantees about tool costs or usage limits.
Who runs it
Bartek Witczak
Software architect and principal engineer with over 15 years building business-critical products across frontend, backend and cloud.
I teach developers in person. Since 2017 I've run in-person workshops, often 2–3 days long, on React, Redux, Next.js, TypeScript and Node.js, including for large companies.
I use AI in my daily engineering work, in legacy and new codebases. I've set up project rules, design systems and my own setups for AI-assisted work, and I build my own products with AI.
AI doesn't replace engineering thinking. It makes good context, architecture, standards and verification matter more.
Questions
FAQ
+Which AI tools do you cover?
We agree the tools during the first call, based on what your team uses or is allowed to use. The method matters more than the tool.
+Do we need a specific tech stack?
Tell me your stack in your first email. I'll say honestly whether I'm a good fit for it.
+Can we work on our own repository?
Yes, that's the preferred way. Before we start, we agree what code and data can be used, with which tools, and who approves it.
+Can we do it remotely?
The workshop is in person. If that isn't possible for you, write anyway and we'll see what makes sense.
+What do we need to prepare?
A project to work on, access to the agreed tools and accounts, and a budget for model usage during the sessions.
+What does it cost?
After a short call I send a proposal with the format, length and price.
+Some of our developers don't believe in AI. Is that a problem?
No. Sceptical developers often ask the best questions. The workshop asks them to test the approach on real work and judge the result.
+Will this make us faster?
I don't promise a speed-up. The aim is clearer plans, better checks and less rework.
Next step
Send me a short email.
A few sentences are enough:
- · Who's on the team, and how experienced they are
- · Your stack and the AI tools you use now
- · What isn't working, or what you'd like to change
I reply, we schedule a short call, and if it fits I send a proposal with format, length and price.
bartek@bartekwitczak.com
Need something different?
Each of these is a separate service, not a next step you're expected to take.
External AI advisor
AI that genuinely improves how your business works.
AI workshops for business teams
Your team knows how to use AI in its everyday work.
AI in software delivery
From AI licences to a shared way of building software.