IT training across Europe - tecnovy makes you future-ready.

Switch to English
E-Learning

tecnovy AI Suite

Cloud AI Architecture

An assistant has to search internal documents, and processing has to stay in Europe. On AWS you choose Amazon Bedrock with a geographic EU inference profile, on Azure an EU Data Zone deployment in Microsoft Foundry, on Google Cloud an EU endpoint of the Gemini Enterprise Agent Platform. Quotas and billing follow different rules on each. You set these options side by side and argue which architecture fits your requirements.
Attendance certificate4.8/5 on ProvenExpert

What this module delivers.

After the tecnovy Cloud AI Architecture training you justify whether an AI workload runs on AWS, Azure or Google Cloud, by services, data locations, quotas and pricing. Cloud and solution architects hand their team a decision others can check.

How the training runs

Method
Comparative architecture exercises, case studies and demos across all three clouds
Basis
tecnovy curriculum on the AI services of AWS, Azure and Google Cloud
Outcome
tecnovy certificate of participation, no examination

Dates & booking

Choose a date that fits

No public dates are scheduled yet.

New dates are published regularly. Join the waitlist or plan an in-house course for your team.

Request in-house training

Fit

Who this module is designed for

Typical roles

  • You design, as a cloud or solution architect, the platforms that AI applications run on.
  • You plan an AI platform for your organisation and decide on its building blocks.
  • You prepare the decision on which cloud provider AI workloads should run with.

Prerequisites

No prerequisites

You can start right away. What you need: a working knowledge of at least one of the three major cloud platforms, experience in architecture work.

Consider instead AI Solution Architecture For those who want to design an AI solution independently of any platform first, from requirements to documented decisions.

Curriculum

Cloud AI Architecture Course in Detail

Five parts in the order of a platform decision: the three providers' services first, then reference architectures, then access and data location, then cost and operation, and finally the choice of provider. Every part sets AWS, Microsoft Azure and Google Cloud side by side.

01AI platforms at a glance

The opening part maps the three providers' AI offerings to each other.

  • Model access: Amazon Bedrock, Foundry Models in Microsoft Foundry, Model Garden in the Gemini Enterprise Agent Platform
  • Development environments such as Amazon SageMaker AI, the Foundry portal and Agent Studio
  • Agent services: Amazon Bedrock AgentCore, Foundry Agent Service, Agent Runtime
  • Where the offerings overlap and where they are cut differently
02Reference architectures

This part builds reference architectures for the typical AI applications in the cloud.

  • GenAI applications that use a model through an interface
  • RAG (retrieval augmented generation): answers based on retrieved company documents
  • Agents in the cloud
  • Serverless (the provider runs the infrastructure), containers or dedicated GPU capacity
03Security and compliance

This part settles who accesses AI services and where data is processed.

  • Identities and rights for applications, agents and users
  • Network isolation and encryption
  • Data residency: in which regions requests are processed and data is stored
  • Provider contracts and requirements from EU law such as the GDPR
04Cost and operation

Here you work out what an AI workload costs and plan how it runs.

  • Billing per token, reserved capacity and batch processing
  • Infrastructure cost alongside model cost
  • Quotas: ceilings on tokens or requests, allocated differently by each provider
  • Monitoring and scaling
05Multi-cloud and choice of provider

The closing part ends in a reasoned decision for one provider or several.

  • Where dependencies on one provider arise
  • How to reduce those dependencies
  • Decision criteria for choosing a provider

Outcome

What you will be able to do afterwards

  1. 01

    You map model access, development environments and agent services of AWS, Azure and Google Cloud to each other.

  2. 02

    You design reference architectures for GenAI applications, RAG and agents on each of the three clouds.

  3. 03

    You choose between serverless, containers and dedicated GPU capacity and justify the choice.

  4. 04

    You define identities, network isolation and encryption for an AI workload.

  5. 05

    You compare the providers' data location options and check them against your organisation's rules.

  6. 06

    You estimate the cost of an AI workload from the pricing models for models and infrastructure.

  7. 07

    You plan monitoring, quotas and scaling for ongoing operation.

  8. 08

    You justify choosing one provider or several and name where dependencies arise.

Offer scope

What is included

Included

  • Comparison overview of the AI services of AWS, Azure and Google CloudIt sets the three providers' services side by side and contains the reference architectures from the training.

Why tecnovy

What you get on top with us

01

No Slideshow, Hands-On!

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.

02

Certificate Showroom

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

Flexible Date Change

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.

04

Attend Twice, Pay Once

You are welcome to attend the training online again within a year as a refresher.

05

Learn from Experts

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.

FAQs

Frequently asked questions

01Which cloud is best for AI?
None of the three is best across the board; the choice follows from your requirements. In the tecnovy Cloud AI Architecture training you weigh where data may be processed, which models are available in your region, how usage is billed, which identity and network structures already exist and how tightly you tie yourself to one provider.
02Can I have AI requests processed in Europe on AWS, Azure and Google Cloud?
Yes, all three providers offer options that keep processing in Europe, each with its own conditions (as of November 2026): geographic EU inference profiles in Amazon Bedrock, Data Zone deployments in Microsoft Foundry, whose EU zone can include EFTA states such as Norway and Switzerland, and European regional or EU endpoints in the Gemini Enterprise Agent Platform. Global options process data worldwide, and not every model is available in every option. Whether a configuration meets your organisation's data protection requirements is your organisation's decision; in the tecnovy Cloud AI Architecture training you compare the options as the basis for it.
03How do AWS, Azure and Google Cloud bill for AI models?
All three offer per-token billing for language models, a token being the unit of text in which models count input and output, and reserved capacity for steady load (as of November 2026): Provisioned Throughput in Amazon Bedrock and in the Gemini Enterprise Agent Platform, provisioned throughput units in Microsoft Foundry. For jobs without time pressure, all three offer batch processing. In the tecnovy Cloud AI Architecture training you estimate a workload's cost from these models, including the infrastructure around the model.
04Why do AI applications hit different limits with each provider?
Because each provider allocates its quotas differently, meaning the ceilings on tokens or requests per minute (as of November 2026). AWS sets them per account for Amazon Bedrock, and they can vary by region and usage history; Microsoft Foundry has been moving to shared per-subscription quotas in stages since May 2026; Google Cloud ties Standard PayGo throughput to your organisation's spend over the last 30 days. In the tecnovy Cloud AI Architecture training you plan scaling around these rules.
05Why do the AI services have different names than in older material?
Because all three providers have renamed their AI services: Vertex AI has been Gemini Enterprise Agent Platform since 22 April 2026, and Azure AI Foundry is now Microsoft Foundry. On 3 December 2024 the ML service Amazon SageMaker became Amazon SageMaker AI, while Amazon SageMaker now names the wider data and AI platform. This page for the tecnovy Cloud AI Architecture training uses the names the providers use in their documentation as of November 2026.
06Do I need accounts on all three clouds?
No, the tecnovy Cloud AI Architecture training does not need your own accounts on AWS, Azure and Google Cloud. The exercises are architecture exercises, and the providers' services are shown in demos.
07How does Cloud AI Architecture differ from AI Solution Architecture?
The tecnovy AI Solution Architecture training designs an AI solution independently of any platform, from requirements and quality goals through architecture patterns to documented decisions. Cloud AI Architecture starts where that ends and compares how such a design is built with the services of AWS, Azure and Google Cloud.
08Is there an examination or a certification?
No, the tecnovy Cloud AI Architecture training ends without an examination and leads to no certification; you receive the tecnovy certificate of participation. If you want a vendor certification for one platform, the tecnovy AI Suite has trainings that prepare for AWS Certified AI Practitioner, Microsoft AI-103 and Google Cloud Certified Generative AI Leader, whose exams are the responsibility of AWS, Microsoft and Google Cloud.

What does your training at tecnovy look like?

Cloud AI Architecture

No dates yet

We will let you know as soon as the next date is scheduled.

Feedback Form
What should we improve?

Tell us what we can do better next time.

Success
Error

© Copyright 2026, tecnovy GmbH, All Rights Reserved