[Br]eaking [Ba]rriers in QA: From Fr[Ai]gile to Agile with AI - A Workshop Recap

February 26, 2025

Topics

This week, our QA Community gathered for a half-day workshop that was as insightful as it was interactive. Codrin Păduraru, Software Tester, delivered an engaging session titled “[Br]eaking [Ba]rriers: From Fr[Ai]gile to Agile with AI.” Building on his popular 10 Things QAs (Should) Know About Agile Scrum workshop from last year, Codrin’s latest presentation dove headfirst into the world of AI - with plenty of exercises, team interactions, and even a few witty AI jokes.

I've always been fascinated by AI — how it works, its capabilities and limitations, and how I can both utilize and learn from it in this rapidly evolving landscape. With AI being a trending topic that captivates individuals and businesses alike, I thought it would be valuable to bring together like-minded people who share this interest. My goal is to stay ahead of technological advancements, exchange knowledge, and possibly even start an AI club. We’ve discussed about how AI works and how it fits into Agile Scrum environments, augmenting testing, the importance of datasets, AI models, biases, ethics, lack of transparency and the future of AI in software testing. I feel like we've only scratched the surface, and I wish we had more time to dive deeper. Hopefully, this sparked some curiosity for a follow-up session! Testers should embrace AI as a tool to enhance quality and accelerate delivery. "AI should augment, not replace, testers by handling repetitive tasks and optimizing coverage. This allows human testers to focus on more complex, creative, and strategic aspects of testing, ultimately improving the overall quality of the software."
- ChatGPT, circa 2025

Codrin Păduraru, Software Tester & Workshop Organizer

A Journey from Fragile to Agile

Codrin’s presentation took us on a transformative journey from traditional, “fragile” quality assurance practices to a robust, agile, AI-enhanced testing environment. He illustrated how the integration of AI with Agile and Scrum practices not only overcomes common testing challenges but also boosts efficiency and fosters continuous improvement.

Agile & Scrum Recap

Agile Philosophy:
Emphasizes iterative development, adaptability, and a customer-centric approach. It’s built on the values of the Agile Manifesto - prioritizing individuals, working software, collaboration, and responsiveness to change.
Scrum Framework:
A lightweight framework that organizes work into Sprints. Codrin recapped key components such as Scrum roles (Scrum Master, Product Owner, Developers), artifacts (Product Backlog, Sprint Backlog, Increment), and events (Sprint Planning, Daily Scrum, Sprint Review, Sprint Retrospective). These elements are designed to foster transparency, frequent inspection, and adaptation.

AI in QA Testing: Welcome to “AI-gile” Testing

In today’s digital era, AI isn’t just a buzzword - it’s transforming the QA landscape. Codrin unpacked several key aspects of AI in testing:

AI Fundamentals:
AI mimics human intelligence using technologies like Machine Learning, Deep Learning, Natural Language Processing, Computer Vision, and Robotics.
How AI Works:
The AI model lifecycle was explained from data collection and training to evaluation, deployment, and continuous improvement. Codrin explored various model types, including traditional ML models, deep learning models, generative AI models, and large language models (LLMs).
Practical Exercises:
The workshop featured hands-on exercises such as training an AI model for image classification, along with creative games like “Prompt My Ride” and “Sprint Game – [Br]eaking [Ba]cklog.”

Challenges, Ethics & Best Practices

No discussion about AI is complete without addressing its hurdles:

Challenges in AI-Powered QA:
Topics such as high initial costs, data quality issues (like biases and algorithmic quirks), and integrating AI with legacy systems were openly discussed.
Ethical Considerations:
The need for transparency, human oversight, and responsible data handling was emphasized - because even AI knows that trust isn’t something you can just code in!
Best Practices:
Codrin advised starting small with clear objectives, choosing the right tools, investing in team training, and maintaining a balance between automation and manual oversight.

Peering into the Future of Agile QA with AI

Codrin also shared his vision for the future:

Emerging Trends:
Look out for augmented intelligent testing, enhanced predictive testing, and even autonomous model generation.
Customization & Collaboration:
Expect AI models to be fine-tuned for industry-specific needs and for improved collaboration between testers and developers.
Enhanced Analytics:
With richer real-time reporting and advanced root cause analysis, our QA processes will soon be smarter than ever.

Sustainability & Broader Impacts

Beyond immediate benefits in testing, the integration of AI has broader implications:

Resource Optimization:
Early testing and smart automation are key to reducing power usage and optimizing server uptime.
Environmental Considerations:
While AI’s computational demands are high, ongoing innovations are addressing challenges such as water usage for cooling data centers and electronic waste management.

What Our Participants Had to Say

Ana Maria Gheorghiu or Anca as she liked to be called, the newest member of our QA Community working on the ChatCGM project, shared her experience:

I really enjoyed the workshop and found the presentation both informative and engaging. Some of the information was new to me, especially the detailed breakdown of AI models (like Generative AI and Large Language Models) and their applications in Agile QA, which I found interesting. The games and interactive exercises, like 'Prompt My Ride' and 'Sprint Game - [Br]eaking [Ba]cklog,' were a great way to apply the concepts and made the learning experience more dynamic. However, I felt that 3 hours wasn’t enough to fully explore all the topics in depth, and extending the time could have allowed for more discussion and hands-on practice. As this was my first workshop at CGM, I was impressed by the level of organization and the quality of the content. Codrin was very knowledgeable and presented the material in a clear and engaging way, which made complex topics easy to understand. The insights I gained will definitely help me in my work, especially in understanding how AI can improve testing efficiency. Overall, it was a valuable experience, and I left with a clear message: AI should complement manual and automated testers, not replace them, by handling repetitive tasks and optimizing coverage.

Ana Maria Gheorghiu, Quality Assurance Professional, Chat CGM Project

Monica Tistu, Manual and Automation Tester on EOS, who joined us last year, added:

The workshop about agile and AI was a good opportunity for me to refresh my knowledge about agile and to learn how I can use AI in my daily work. It was also an opportunity to meet other colleagues from the testing community and to remember how cool it is to work from the office. The workshop combined the useful with the pleasant, we had team games, Codrin, through his charm, managed to capture the attention of his colleagues and convey quality and useful information. Participating in Codrin's workshop was the start of paying more attention to AI and understanding how it is designed and how it can help me in testing.

Monica Tistu, Quality Assurance Professional, EOS Project

Looking Ahead

The QA Community is growing rapidly, and they have big plans for this year - including a new surprise that’s already in the works. We encourage all CGMers to organize and deliver workshops and presentations on topics they are passionate about, like Codrin did. With AI playing a pivotal role in shaping the future of eHealth and in delivering exceptional healthcare services every single day, there’s no better time to get involved.

Let’s continue to break barriers, one sprint at a time, and remember: while AI might help us automate our tests, it’s the human touch that truly makes our QA practices robust, resilient, and a little bit fun.