Skip to content

Agentic Engineering training

Launch Agentic Engineering in days.

Your whole team builds real agents and quality gates on its own stack and code. People leave with a shared practice and artifacts they can use on Monday.

Two days | real code | German or English

Bene speaking during an engineering workshop

One delivery system

We train whole teams. Not roles

In agentic engineering, product people write specifications agents consume, developers shape product decisions, and designers prototype with AI. The team trains together because the team delivers together.

Context engineering foundations

Multi-layer context architecture, project rules, MCPs, skills, and vendor-independent setups.

Agents across the SDLC

Agents support refinement, planning, implementation, review, testing, documentation, and legacy migration.

Team practices

Pairing with agents, review protocols, knowledge sharing, and shared standards replace isolated experimentation.

Architecture and quality

Domain boundaries, API contracts, test strategy, and quality gates hold up when agents write code.

The shift

From individual tools to coordinated practices

Individual tools

Your team has AI tools. Adoption is still fragmented.

  • Senior engineers dismiss AI as not real engineering.
  • Everyone uses the same tools differently.
  • Early adopters create output nobody else maintains.
  • Team velocity does not improve consistently.
  • Quality concerns and fear of craft loss remain.

You lose cohesion, confidence, and the value of the tool investment.

Coordinated practices

The team builds shared capability on real work.

  • Skeptics learn from evidence and peers.
  • Agents follow shared context and standards.
  • AI becomes part of refinement, planning, and review.
  • Configurations and practices live in the repository.
  • The team can keep improving without the trainer.

Capability moves from individuals into the delivery system.

The journey

Learning becomes delivery immediately

  1. 01

    Pre-training assessment

    Current AI use, pain points, goals, technical constraints, and the codebase shape the session.

  2. 02

    Two hands-on days

    Four modules move from foundations into your repository and your real development lifecycle.

  3. 03

    Deploy immediately

    Project rules, agents, skills, quality gates, and the starter kit leave with the team.

  4. 04

    Follow-up and pilot

    Q&A support and developer community access keep the practice alive while the strongest value stream becomes the pilot candidate.

The transfer ends in the team's repository. Project rules, agents, skills, and quality gates remain available after the two training days.

See it first

A live demonstration before you commit

No slides-only pitch. See context engineering, agents, and quality gates work on a reference codebase or yours.

Free 60-minute live demo

A practitioner runs the real setup in a live development environment and answers the questions your team will have.

Ask about the live demo

The people in the room

Workshops are built around a real team

The format brings people with different responsibilities into one shared engineering practice.

Björn, Tereza, Fabian, Denis, Tim, Bene, and Stefan together after a workshop

Björn, Tereza, Fabian, Denis, Tim, Bene, and Stefan after an engineering workshop.

The group around a session can include facilitators, practitioners, and people from the wider network. The training itself is led by experienced engineering practitioners.

Teams share what changed after training

Dr. Jan Henrik Ziegeldorf
CTO & Co-Founder aedifion
The workshop was inspiring and pushed me to think big. The end-to-end migration pipeline example really stayed with me. We're using Claude Code far more aggressively now.

Training

Jan Tegtmeier
Managing Director Tegtmeier Internet Solutions
Since the workshop, everyone here develops agentically. Some still work with agentic support, others haven't written a single line of code since. I'd recommend a workshop with you to any developer.

Pilot

Michael Mohring
Head of Product Development adesso insurance solutions
We work with AI much more deliberately now. 'Agent first' has become our reflex: do we need to do this ourselves, or can an AI agent handle it? The collaboration took the fear away from our developers.

Rollout

Pilot

We went from 2 engineers playing with agentic coding to 20 using it daily. We created 17 skills and 9 agents available to every engineer. Skeptics now embrace it, with a Slack channel dedicated to sharing learnings and complaints when Claude goes down. We're targeting daily sprints across the full product development lifecycle by summer.
Tom Hibbert
VP Engineering Moonfare

Funding

German companies may qualify for training support

Eligibility and coverage under the Qualifizierungschancengesetz depend on the company and participants. We can look at the route with you. We promise no funding outcome.

Check eligibility early

Bring company size, participant roles, employment context, and the planned training scope into the first conversation.

Discuss eligibility

Your trainers

Practitioners who build production systems

Every workshop is led by people with architecture and engineering experience, not by tool trainers.

Selected trainers

Training formats, logistics, and next steps

How is Catalyst different from Team Training?

Catalyst creates awareness and buy-in through a live demonstration. It does not build the full team capability. The two-day Team Training is the hands-on engagement that produces shared practices and repository artifacts.

Is this only for developers?

No. Product managers, product owners, designers, engineering leaders, and developers can train together. The curriculum is adapted to the roles present while preserving one shared delivery system.

Will the training become outdated as tools change?

The core is context engineering, working with agents, team practices, architecture, and quality. Tools change. Those principles and the infrastructure in your repository stay useful.

Do you work with distributed teams?

Yes. We prefer on-site for the hands-on days, but remote works with cameras on, shared exercises, and direct access to the working repository.

How do you adapt to our stack?

The implementation details follow your languages, frameworks, architecture, development lifecycle, and security constraints. The team works on its own code rather than a generic tutorial project.

What team size works?

Recommended cohort size is 8-12. Smaller teams get an intensive format. Larger cohorts change the delivery plan, and groups above 20 are scoped explicitly.

Can we start with one team?

Yes. One real team is the normal starting point. Training creates a candidate pilot team and evidence for deciding how to expand.

What happens after training?

The team keeps the artifacts. Follow-up support and community access for the participating developers keep the practice moving. The next commercial step is usually a Factory pilot in one real value stream.

We already use GitHub Copilot. Why train?

Autocomplete is only one part of the lifecycle. Training builds the context infrastructure and team practices that make Copilot, Claude, Cursor, and future tools work consistently across requirements, planning, review, testing, documentation, and delivery.

Next steps

Put the practices to work

Build the capability, then prove it in a pilot

Tell us where delivery gets stuck. We will work out whether training, the readiness assessment, or another entry point fits your situation.