Principle 01
Roles instead of prompts
Agents work in fixed roles: requirements, planning, design, code and verification. Every task passes through the same stations.
AI consulting and agentic engineering with governance
AI works
for you.
You stay
in control
I introduce AI in companies and development teams: training, a protected AI environment, agents and software development with agents. With fixed rules, human approval and prices that are settled beforehand.
How a change flows · demoWaiting for approval
- 01RequirementAgent
- 02PlanAgent
- 03DesignAgent
- 04CodeAgent
- 05VerificationAgent
- 06ApprovalHuman
- 07ReleaseAgent
No merge without approval. Gates before it: Definition of Ready, at least 80% test coverage, security scan.
01 Services
Three ways in, depending on where you stand.
Each one starts with a small, clearly defined step. What it costs is stated next to it.
How do we use AI safely and to real effect?
Introduce AI
For management, IT and business units: train the team, set up a protected AI environment, hand recurring work to agents.
- AI literacy training
- €1,900
- AI policy & governance check
- €3,900
- Agent Start
- €6,900
Introduce AI
How do we ship faster without losing control?
Build software & cloud
For CTOs and engineering leads: introduce development with agents, review architecture, cloud and code, build products.
- Agentic engineering assessment
- €7,900
- Cloud architecture assessment
- from €5,900
- Architecture and code review
- €6,900
Software & Cloud
Who leads the project technically?
Interim leadership
For companies whose project needs technical leadership from outside: architect, technical project lead, product owner, tech lead or interim CTO.
- Framework contract
- by effort
- Duration
- typically 3 to 12 months
- Location
- remote and on site
Interim leadership
All prices exclude VAT. The offer is aimed exclusively at businesses.
02 Approach
Agents in fixed roles. Gates at every step. People accountable.
This is how I build software myself: every task passes through the same stations. Whatever fails a gate does not move on.
Principle 02
Gates instead of hope
Tests, type checks and security scans run automatically. Whatever fails a gate does not move on.
Principle 03
Approval by people
No merge without review and approval. Accountability stays with your team.
03 Guide
Which entry point fits you?
Tick what applies to you. You see at once which entry point fits.
Recommendation
Tick what applies. The recommendation appears here.
The evaluation stays in your browser; nothing is transmitted.
04 About me
Architect, tech lead and a daily user myself.
I am Ludwig Oswald, a cloud and solution architect based in Munich. For more than ten years I have been responsible for architecture, teams and security in software projects, including automotive, mobility and medical technology.
I am currently CTO of a start-up whose development runs end to end with AI agents: fixed roles, a versioned rule set, gates in the pipeline and human approval. What I introduce for you, I use myself.
- Architecture
- Cloud and solution architecture, platforms on AWS, Azure and GCP
- AI engineering
- AI agents, MCP integration and software development with agents
- Compliance
- DevSecOps, including BSI C5 and ISO 27001
- Certified
- iSAQB CPSA-F, PSM I, PSPO I, ITIL
- Location
- Munich, remote and on site
05 Contact
30 minutes to find out whether it fits.
In the intro call we look at your situation: team size, tools, and what has changed since AI tools were introduced. Afterwards you receive an offer within one working day, or a clear recommendation not to proceed.