Coding agents now write whole features, and they respect only the architectural decisions that were given to them and that something actually enforces. That moves your own work from writing the implementation towards orchestrating it. AGENTA runs through seven curriculum chapters: from the fundamentals through decision-making, knowledge delivery and guardrails to governance and the shape architecture work takes next.
AI-assisted development produces architectural drift faster than review cycles catch it. The tecnovy AGENTA training shows you how to supply agents with architectural knowledge and hold them inside guardrails, so your team keeps the speed without losing the structure.
Practice in six of the seven chapters, 315 of 900 minutes
Official iSAQB curriculum AGENTA 2026.1-rev5, English only
30 credit points: 20 methodical, 10 technical, no course exam.
Dates & booking
4 dates
Sessions with this symbol offer up to 25% group discount. Click “Details & Registration” to learn more.
2,510 $−377 $2,133 $
2,510 $
excl. VAT2,510 $
excl. VAT2,510 $−251 $2,259 $
No dates match this selection.
Dates are still available. Reset the filters to see them.
Fit
You answer for architecture decisions in a team that works with coding agents.
You decide which architectural knowledge agents receive, and in what form.
You have to make AI use in your company evidenced and compliant.
No prerequisites
You can start right away. What you need: hands-on practice in software development or software architecture, a working grasp of repositories, build pipelines and everyday development tooling. Helpful: some first-hand experience with AI assistants or coding agents.
Consider instead SWARC4AI Covers the architecture of AI systems themselves, from data pipelines through model serving to drift and RAG. AGENTA instead puts agents to work as a tool of architecture work.
Curriculum
Curriculum 2026.1-rev5 splits AGENTA into seven chapters across 585 teaching and 315 practice minutes. It starts at the fundamentals of LLMs and agents, runs through architectural decisions, knowledge delivery and guardrails to the extraction of architecture information, and closes at governance and the changing role of the architect.
The curriculum uses this chapter to bring along participants with no prior AI experience.
This is where a decision becomes an experiment instead of a meeting.
The largest block: what an agent knows decides what it builds.
Knowledge alone is not enough; this chapter builds the enforcement.
The direction reverses here: from the system back to the documentation.
This chapter settles what your company has to be able to evidence.
To close, the subject is your own role and your team's.
Outcome
You separate AI-assisted, agentic and vibe coding from each other and decide per task between a human, a classic tool and an agent.
You frame architectural decisions as testable hypotheses and validate them with spikes, prototypes and side-by-side variant comparisons.
You design agent-facing representations of architectural knowledge, from ADRs through context files to retrieval indexes.
You choose delivery mechanisms for that knowledge and judge where the Model Context Protocol genuinely helps a team.
You design preventive and detective controls that enforce architectural decisions on AI-generated work.
You recover architectural information from code, tests and configuration, and generate current architecture views from it.
You assess prompt injection, tool misuse and sandboxing against the OWASP LLM Top 10.
You map the EU AI Act, copyleft licences and data-protection roles onto AI use in your own company.
The tecnovy certificate of participation records your attendance of the AGENTA training, not a passed examination.
Open the Certificate Showroom tecnovy →
Why tecnovy
01
We are an officially accredited Training Provider of the International Software Architecture Qualification Board.
02
Get your certificate of participation and, if you have one, add your exam certificate from E-Learning. Fully automated, beautifully designed. Just for you, only at tecnovy.
03
Promised: no PowerPoint marathon. We work in groups, tie theory to practice, and you get real project examples from our experienced trainers plus the exchange with like-minded people.
04
You are welcome to attend the training online again within a year as a refresher.
05
We always guarantee you the use of didactically and methodically first-class qualified trainers who draw their knowledge from training experience as well as professional practical and project experience.
06
If you are not able to attend the course, you can rebook your training free of charge up to one week before the start of the training.
FAQs
No. The iSAQB curriculum for the AGENTA module lists practical experience in software development or software architecture as its prerequisite, not a certification. For the later CPSA-A certification, CPSA-F is mandatory.
No. There is no examination for the AGENTA module. The tecnovy AGENTA training earns you 30 credit points and a certificate of participation; the assessment comes later on the path to CPSA-A certification, and it covers all modules together.
CPSA-A certification consists of a written assignment. Two examiners recognised by the iSAQB assess it, and you then defend it in discussion with them. You can take it in German or English, and tecnovy is glad to help if you have questions about the process.
The tecnovy AGENTA training runs for three days, and the iSAQB curriculum 2026.1-rev5 names three days explicitly as its minimum. Of the 900 minutes, 315 are exercises, more than a third; the heaviest chapter is providing architectural knowledge to agents, at 150 teaching minutes.
The direction: AGENTA treats AI as a tool of architecture work, where agents write the code, you supply them with architectural knowledge, hold them inside guardrails and settle governance and traceability. SWARC4AI treats AI as the subject of architecture work: data pipelines, model serving, drift, RAG and EU AI Act classification for systems that contain AI themselves. If you build AI systems, book SWARC4AI at tecnovy; if you develop with agents, book AGENTA.
No. Prompt and context engineering is one of six learning goals in the opening chapter, which runs 60 teaching minutes in total. The weight of the tecnovy AGENTA training then falls on architectural knowledge for agents, guardrails and governance.
No, the iSAQB curriculum prescribes no product. It describes the agentic loop, the autonomy spectrum and orchestration patterns, which is the part that survives a change of vendor. First-hand experience with any coding agent will still help you in the tecnovy AGENTA training.
No. The tecnovy AGENTA training places the EU AI Act, copyleft licences and the roles in data protection in context, so that you recognise when to involve a data protection officer or information security and how to document the decision. The legal assessment of a specific case is not part of the course.
AGENTAAGENTA - Architecture for Agentic Software Engineering Contexts
Choose a public date or plan the course for your team.
Tell us what we can do better next time.
© Copyright 2026, tecnovy GmbH, All Rights Reserved
