"Accelerate your Testing with AI"
Compliance to the Business - Methods
Goals:
- Master the various AI model querying techniques for software testing activities
- Accelerate test analysis/design, test data creation, test automation and test maintenance with AI.
- Understand and assess the risks of using AI to speed up software testing
Profile targeted
- Quality Engineer, QA Manager
- Tester, Consultant Testing
- Developer, Test Automator
- Product Owner, Project Manager,
- Business Analyst
Mandatory Requirements
None
PRESENTATION
Discover our program:
1. Generative AI for software testing: Introduction
- Different types of AI: machine learning, deep learning and generative AI
- What generative AI brings to software testing
- The main techniques for using generative AI for testing
2. Requesting a large language model for testing: how to get good results
- Prompting techniques
- Use cases with practical exercises :
- Test analysis and design
- Refactoring / improving existing test cases
- Generate test data
- Generate and maintain automated tests
- Analyze fault reports and summarize problems detected
- Summary: best practices for querying an AI model
3. Generative AI integrated into test tools
- How AI is infusing testing tools
- Practical experience:
- Regression tests
- Resuming a manual test sequence
- Assistance in correcting test code
4. Managing the risks of generative AI
- My AI is hallucinating: practical examples
- Mon IA ne protège pas mes données : testez-le
- My AI is biased: test it
- My AI costs too much and its energy impact is too high
=> Solutions to these problems
5. Advanced techniques for using generative AI in software testing
- Using our project data with generative AI
- Guiding reasoning
- Provide examples
- Query string
- AI agent tester: your assistant for more complex tasks
- How do you adapt an AI model to test tasks?
6. Synthesis: what we've learned
- How do I go about it in practice?
- I choose my generative AI model for testing
Tools:
- GPT 3.5 and GPT 4 (OpenAI)
- Mistral 7B and Mixtral (Mistral)
- Gravity (Smartesting)
- Llama and CodeLLama (Meta)
- Gemini (Google)
Training will adapt to technological developments, and this list will evolve according to what is most relevant.
Teaching methods
35% theory / 65% practice
evaluation methods
None