dominikfretz.com

What It Is

Dominik Fretz is not a traditional SaaS tool, but a set of training, workshop, and consulting services centered on Agentic AI Engineering and Claude Code. The site’s core argument is that AI is driving the cost of “writing code” close to zero, but building maintainable production software remains difficult. As a result, the service does not focus on simply teaching teams to use ChatGPT, Copilot, or Claude to generate code, but on helping them establish a reusable AI-assisted engineering methodology.

Core Capabilities and Methodology

Its main methodology is Spec → Phases → Plan → Implement: first define the requirements specification, then break the work into phases, have AI generate an implementation plan, and finally implement under human constraints, testing, and quality gates. The content repeatedly mentions tools such as Claude Code, Cursor, and OpenAI Codex, with Claude Code as the primary tool in its stack. Compared with “vibe coding,” where an app is generated from a single prompt, this approach focuses more on code that can be reviewed, maintained, and shipped to production, making it more suitable for engineering teams with production-quality requirements.

Pricing and Delivery Formats

The site lists three types of engagements: a half-day Claude Code Masterclass for individual developers and small teams; a full-day Team Transformation Workshop for engineering teams of 5–20 people; and a multi-day Corporate Training Program for organizations with more than 20 people. The page does not publicly disclose specific pricing, payment methods, or refund policies. Interested buyers need to book a Discovery Call for further discussion, which is not ideal for procurement budgeting.

Pros and Cons

The strengths are its focused positioning, clear methodology, and the instructor’s 20+ years of production software engineering experience, along with his self-described role as an Anthropic Claude Community Ambassador. The content covers PRD-driven development, prompting for complex features, code review, team standards, shared prompt libraries, and ongoing coaching, all of which can be practically valuable for engineering managers. The drawbacks are the limited number of public case studies and quantified results. There is also no clear explanation of data privacy, corporate confidentiality, compliance, or post-training support SLAs. In addition, the service is highly dependent on delivery by an individual expert rather than being a self-service software platform.

Who It’s For and Access from China

It is better suited to engineering teams that are already using tools such as Claude Code, Cursor, or Copilot but are concerned about technical debt, code quality, and team collaboration getting out of control. It is less suitable for users who simply want an online AI coding product. The text does not state the access situation from mainland China. However, the overseas tools it depends on, such as Claude Code and OpenAI, may face network and payment restrictions. Domestic teams in China may also consider alternatives such as 通义灵码, 腾讯云 AI 代码助手, and MarsCode, as well as localized training resources.