Agentic systems and architecture decisions
Deterministic workflows, routing, parallelization and plan-execute loops.
Build useful, reliable agents
Agentic AI Engineering
Design useful, reliable agents ready for real-world use.
Design agents that use tools and skills, connect through MCP, collaborate through A2A and remain observable and controllable.
First sessions planned from 14 August 202620 participants maximum per groupCentral European Time (CET/CEST)
Artificial IntelligenceLIVE COURSE
20 h
of group sessions
Advanced
level
FR + EN
your choice of language
01 — OUTCOME
What you’ll be able to do
Choose between a deterministic workflow and an autonomous agent
Design robust tools and reusable Agent Skills
Manage state, memory, persistence and context
Connect tools and agents with MCP and A2A
Evaluate, secure, observe and deploy an agentic system
02 — CURRICULUM
A clear, progressive learning path
Detailed content may evolve to stay aligned with current tools and practices.
Tools, structured actions and Agent Skills
Function calling, schemas, error recovery, permissions and SKILL.md-based skills.
Context, state and memory
Runtime state, short- and long-term memory, persistence and user isolation.
MCP and A2A protocols
Tools and data through MCP; agent discovery, delegation and collaboration through A2A.
Orchestration and human control
Managers, specialists, handoffs, subagents, approvals and least privilege.
Evaluation, observability and production
Traces, evals, regressions, guardrails, cost, latency, resilience and deployment.
Project & deliverables
- A reusable Agent Skill with an MCP integration
- A multi-tool agent with persistent state and human approval
- A containerized agentic application with traces and an evaluation suite
Prerequisites
Intermediate Python, HTTP/JSON APIs, Git and initial experience with an LLM API. No prior multi-agent experience is required.
03 — FORMAT
Live guidance, practice and feedback.
Group sessions
Demonstrations, exercises, reviews and Q&A over video calls.
Practical deliverables
Guided outputs that apply the skills to realistic scenarios.
Discord
A private space for announcements, questions and peer support.
05 — ENROLMENT
One clear plan, delivered live with the group.
One-time payment. Software subscriptions, API credits and cloud services are not included.
Voluntary satisfaction guarantee
Join with confidence: request a full refund before the second group session begins.
Cohort
€899 TTC
The complete course, delivered live with your group.
- 20 h of live group training
- Exercises and guided deliverables
- Supporting resources
- Private Discord
- Certificate of completion
Loading your preferences…
Full refund on request before the 2nd group session.
Your language and time-slot preferences are saved before payment. The shared dates will then be confirmed by email.
06 — FAQ
Before joining the cohort
How is the course delivered?
The course includes 20 hours of live group sessions, exercises, practical deliverables and supporting resources.
When will I receive the schedule?
Before paying, you choose the course language and every time slot that could work for you. Your group’s 10 shared dates will then be confirmed by email.
What does the Cohort plan include?
The plan includes 20 hours of live group training, guided exercises and deliverables, supporting resources, access to the private Discord and a certificate of completion.
What certificate will I receive?
At the end of the course, you will receive a Genial AI Academy certificate of completion confirming that you completed the training. It is separate from a professional certification.
Is the course recorded?
The initial format is based on live sessions. Access to a complete recorded course library is not promised.
Can I pay by card or PayPal?
Yes. One-time payment is available by card through Stripe or with PayPal.
How does the guarantee work?
You may request a full refund before the second group session begins, without giving a reason.
Ready to begin?
A practical path, live interaction and a community to grow with.
Check my availability