Workshop · Last updated July 31, 2026
Claude Code Workshop for Your Engineering Team
A hands-on Claude Code workshop for engineering teams, run remotely over one day (1,500 EUR) or three days (5,000 EUR). Your engineers do not watch a demo. They audit your current setup, build a working sub-agent for one of your real workflows, and leave with a roadmap for rolling it out across the team. Delivered by our CTO.
Who delivers it
The workshop is delivered by Nicu Constantin, CTO at DeviDevs. He builds and secures the agent systems we ship, and he teaches: he is the assigned practice tutor for our UTCB cohorts, most recently a group of 18 students in summer 2026, six live sessions plus individual project work, closing with a colloquium in July 2026. He also advises the university on course content.
You get the engineer who builds this for a living, not a trainer reading someone else's deck and not a subcontractor. Nicu has written more than 90 technical articles on our blog, on agent architecture, prompt injection defense, and AI security. If your team wants to argue with him about how a hook should fail, that is the point of the day.
What your team walks away with
Four concrete outputs: an honest audit of your current AI dev tooling maturity, a working sub-agent your engineers built during the session on one of your real workflows, the production patterns behind it (sub-agents, hooks, MCP, skills), and a written roadmap for scaling Claude Code across the team.
- An audit of where you are now. What your team already does well with AI tooling, and where the time is actually going.
- A sub-agent that runs. Built during the workshop, on a workflow you picked, on your stack. Not a toy example.
- The production patterns. Sub-agents, hooks as enforcement, MCP integrations, and reusable skills. Why each one exists and when it is the wrong tool.
- A rollout roadmap. A written document your team can act on after we leave, not a feeling that it was a nice day.
One day or three days
The one-day format covers the core: setup, the patterns, and one working sub-agent. The three-day format adds depth on reliability and enforcement, integrations into your own systems, and a second agent your team builds with less help, which is where the learning actually lands.
One day · 1,500 EUR
Remote, hands-on, your team's stack
- Audit of your current AI dev tooling maturity
- How agentic workflows are structured: sub-agents, hooks, MCP, skills
- Your engineers build one working sub-agent for a real workflow
- Roadmap document for scaling Claude Code across the team
Three days · 5,000 EUR
Everything above, plus the parts that keep it alive
- Reliability in practice: retries, deduplication, circuit breakers, what to do when a model returns garbage
- Hooks as enforcement, so team rules live in code instead of in a document nobody reads
- MCP integrations into your own systems: issue tracker, mail, internal APIs, data
- A second agent, built by your team with less help from us
- Review of what they built, with the failure modes named
Who this is for, and who it is not for
This is for engineering teams that already write code and want to use AI agents seriously in their own workflow. It is not an introduction to AI for non-technical staff, and it is not a sales session for our build work.
- Good fit: a team of 3 to 12 engineers, with a real workflow they want to automate and access to their own systems during the session.
- Poor fit: a general AI awareness session for a mixed audience. We will tell you so instead of taking the booking.
- Not a pitch. If your team can do the work themselves after the workshop, that is the correct outcome and we are fine with it.
Price
1,500 EUR for the one-day workshop, 5,000 EUR for the three-day format, remote. The first workshop we run for a company is free in exchange for a case study we can publish.
The price covers preparation on your stack, delivery, and the roadmap document. Travel is quoted separately if you want it on site. There is no per-seat fee: the price is for the team.
Book the workshop
Tell us the size of your team and one workflow you would like them to automate during the session. We will confirm whether it fits in a day or needs three.