Question 1

What is AMP?

Accepted Answer

AMP (Axini Modeling Platform) is a model-based testing platform developed by Axini. It is a platform that allows you to thoroughly test most systems in a fully automated way based on models. Instead of manually writing test scripts, you model the desired behavior of your system and AMP automatically generates test cases based on this model.

Question 2

What is the core functionality of the Axini Modeling Platform (AMP)?

Accepted Answer

Model-based testing: You create a formal model that describes the expected behavior of your system Automatic test generation: AMP automatically generates test cases from this model Intelligent test execution: The platform executes tests and verifies whether the actual system behavior matches the model Real-time feedback: During testing you receive immediate feedback on deviations between the model and reality

Question 3

What is model-based testing (MBT)?

Accepted Answer

Model-based testing is an approach where a formal model of the system to be tested is created. This model is then used to automatically generate, execute and analyze tests. The model is a structured representation of the system, often in the form of diagrams or formal descriptions.

Question 4

What is the difference between AMP (Axini Modeling Platform) and MBT (Model Based Testing)?

Accepted Answer

AMP (Axini Modeling Platform) is the software platform that executes models, automatically generates tests and drives test execution. MBT (Model Based Testing) is the testing method where a formal model of the system is built to systematically derive test cases. You can think of MBT as the approach, while AMP is the tool that executes this approach. AMP makes MBT fully automated.

Question 5

What is the difference between traditional testing and testing with Axini?

Accepted Answer

Test design: In traditional testing, you manually write test cases based on requirements and experience. With MBT via Axini, you model the desired system behavior and AMP automatically generates test cases from this model. Test coverage: Traditional tests only cover scenarios a tester thinks of, creating blind spots. AMP systematically explores all possible paths and states in the model, leading to significantly higher test coverage. Maintenance: With traditional testing, every system change requires manually updating all affected test scripts. With MBT, you only update the model and tests are automatically regenerated. Repeatability: Manual tests are prone to human errors and inconsistencies. AMP executes tests automatically and consistently, every single time. Scalability: The number of traditional tests grows linearly with complexity, becoming unmanageable. MBT scales efficiently as one model can generate thousands of test cases. Feedback: Traditional testing provides feedback late in the process. AMP offers real-time insight into deviations between the model and actual system behavior. Knowledge retention: In traditional testing, test knowledge often resides in individual testers' heads. The Axini model serves as living documentation that preserves and shares knowledge.

Question 6

Is testing with AMP suitable for every type of project?

Accepted Answer

MBT is particularly useful for complex systems with many interactions and states. For simpler projects, traditional testing may suffice, but even then MBT can add value through systematic test generation and coverage.

Question 7

What is the AI quality crisis?

Accepted Answer

The AI quality crisis is the growing risk that AI systems increasingly reinforce themselves without sufficient human oversight. AI models are increasingly trained on AI-generated data, causing errors to accumulate. In code, this manifests as subtle bugs that are hard to detect, because the model repeats patterns without true understanding. AI-driven test and review tools have the same blind spots as the generator, causing bad code to be approved. The end result: applications that seemingly work, but fail under edge cases or contain security vulnerabilities.

Question 8

Are there examples of what can go wrong with AI-generated code?

Accepted Answer

2025: An autonomous coding agent wiped an entire production database and then generated fake accounts and false logs to cover up the mistake. 2025: A major tech company patched over 1,100 security vulnerabilities in a single year, partly because up to 40% of code was AI-generated. 2025: The number of web application outages rose significantly, with AI-generated code repeatedly cited as a contributing factor. 2025: More code per developer thanks to AI, but incidents per deployment rose by nearly 25% and failure rate by 30%. January 2026: Newer AI models produce code that doesn't crash, but silently disables security checks or creates fake output, far more dangerous than a regular error.

Question 9

How does Axini help with the AI quality crisis?

Accepted Answer

Independent validation: Test generation is based on formal models, completely decoupled from the AI that wrote the code. The validator has different blind spots than the generator. Breaking the feedback loop: A human-specified formal model serves as the source of truth, not the AI output itself. This way, AI no longer validates itself. Fail early, not silently in production: Axini systematically covers the full behavioral domain of an application, including edge cases where AI-generated code most often silently fails. Scalable with AI development speed: Because Axini's AMP is scalable with AI development speed. Test generation is automated, allowing quality assurance to keep up with the growing speed and volume of AI-generated code. Proven where it matters: It has been deployed in critical sectors such as rail, insurance and life sciences. These are environments where production failures are unacceptable, just as with AI-generated code.

Question 10

Isn't creating a test model just double work?

Accepted Answer

Quite the opposite! A test model is a smart investment that pays for itself quickly. You capture knowledge once in a clear, formal model and that model then automatically generates hundreds of tests for you. This saves enormous amounts of time on manual testing, significantly improves test coverage and catches errors much earlier. Moreover, the model evolves with your system, making maintenance simpler rather than more complex.

Question 11

Does using a test model increase the amount of work?

Accepted Answer

Initially, creating a test model may seem like more work, but this is offset by the benefits of automatically generated tests, better test coverage and the ease of maintaining and updating tests.

Question 12

How long does it take to create a test model compared to traditional tests?

Accepted Answer

The time required to create a test model may initially be longer than writing traditional tests, but this is offset by the benefits of automatic test generation and ease of maintenance.

Question 13

How can MBT help reduce the total testing effort?

Accepted Answer

Although setting up a model takes time, MBT can reduce the total testing effort through reusable, automatic test generation. This saves time and resources when writing individual test cases and performing regression tests.

Question 14

Why should I use a test model if I already have traditional tests?

Accepted Answer

Using test models can help improve the consistency and completeness of tests, identify errors earlier in the development process and reduce long-term maintenance costs.

Question 15

Can MBT replace traditional tests?

Accepted Answer

MBT can complement or even replace traditional tests, depending on the complexity and requirements of the project. It offers a systematic way to generate tests that can supplement or improve manual tests.

Question 16

Can I reuse existing traditional tests in a test model?

Accepted Answer

Yes, elements of existing tests can often be integrated into or converted to the test model, reducing the initial effort.

Question 17

What are the long-term benefits of using a test model?

Accepted Answer

Improved test coverage Easier maintenance and updates Automatic test generation Greater reliability of the test system Faster identification of system errors Lower maintenance costs

Question 18

How can a test model help reduce errors in testing?

Accepted Answer

A test model enforces a systematic and thorough approach, making it more likely that all aspects of the system are tested and errors are detected early.

Question 19

How does a test model ensure better test coverage?

Accepted Answer

Test models enforce a systematic approach that helps identify all possible scenarios and edge cases. By using a formal model, all possible states and transitions of the system can be described, leading to more comprehensive test coverage than manual tests.

Question 20

How does a test model ensure consistency in testing?

Accepted Answer

A test model ensures consistency by providing a standardized and repeatable method for test generation, reducing human errors and variability.

Question 21

How does modeling contribute to system quality?

Accepted Answer

Modeling provides a structured and systematic approach, helping to identify and fix errors early in the development process, thereby improving the overall quality of the system.

Question 22

How does a test model help with test maintenance?

Accepted Answer

Test models are usually easier to maintain and update than traditional tests. Changes to the system can easily be made in the model, which automatically updates the necessary tests.

Question 23

What are the cost savings of a test model in the long term?

Accepted Answer

Less manual testing effort required Fewer errors in production Easier maintenance and updates of tests Improved testing efficiency

Question 24

What are the initial costs of MBT?

Accepted Answer

The initial costs of MBT include the time and resources needed to create the model and to train team members in the use of MBT tools and methods.

Question 25

How can modeling improve communication within a team?

Accepted Answer

A shared model provides a common understanding of the system, improving communication and collaboration between developers, testers and other stakeholders. It provides a formal description that helps prevent misunderstandings.

Question 26

Can modeling help identify system requirements?

Accepted Answer

Yes, by creating a formal model, system requirements can be more clearly defined and verified, helping to prevent misunderstandings and missing requirements.

Question 27

Is specific training needed to work with test models?

Accepted Answer

Yes, some training may be needed to work effectively with test models, but the investment in training pays off in the form of improved testing efficiency and reliability.

Question 28

Is there a learning curve associated with MBT?

Accepted Answer

Yes, there is a learning curve. It takes time and effort to become familiar with creating and using models, but the investments in training and knowledge pay off in the form of more efficient and effective testing processes.

Question 29

Are there tools available to support MBT?

Accepted Answer

Yes, there are various tools available that support modeling and model-based testing, such as Axini, which allows models to be created and tests to be automatically generated.

Frequently AskedQuestions

General

AI

Effort and Efficiency

Benefits and Added Value

Quality and Test Coverage

Maintenance and Costs

Collaboration and Communication

Training and Tools

Still have questions?