How To test From Quality Engineering to AI Trust - Mary Pombares
Categories: Podcasts , How To Test This?
Quality engineering faces challenges and opportunities in AI-driven testing, requiring trustworthy, unbiased systems and strategic implementation. Human expertise remains vital, alongside addressing systemic issues like diversity and empowering QA leaders for sustainable success.
How To Test This?
interview episodes where Mamadou N’diaye talks with with software testing experts
- https://podcasters.spotify.com/pod/show/spidey1944
- https://www.youtube.com/@HowToTestThis
- https://www.linkedin.com/in/mamadou-ndiaye-consultant/
Episode Details
- Show Notes: https://podcasters.spotify.com/pod/show/spidey1944/episodes/How-To-test-From-Quality-Engineering-to-AI-Trust---Mary-Pombares-e3ovoqu
- Published: 2026-09-17T10:46:53Z
- Duration: 00:27:41
- Author: Mamadou N’diaye
Overview
The podcast discusses key challenges and advancements in quality engineering (QE), emphasizing the importance of building a strong quality culture where accountability is shared across teams. A major focus is on the role of AI in testing, including the need to ensure AI systems are trustworthy, unbiased, secure, and free from hallucinations. The discussion highlights that while AI can accelerate test case generation and reduce manual effort, success depends on starting with clear business problems rather than adopting tools for their own sake. Only a small fraction of AI initiatives succeed in reaching production, underscoring the need for strategic implementation, risk assessment, and human oversight.
Further topics include the evolving role of QA professionals in an AI-driven landscape, where automation and AI assist in tasks like regression testing and risk analysis, allowing testers to focus on higher-value work. Despite advancements, human expertise remains critical for architectural understanding, domain knowledge, and ensuring ethical AI deployment. The conversation also addresses systemic issues in tech, such as lack of diversity and underempowerment of QA leaders, calling for inclusive leadership and organizational support. Additionally, the importance of continuous learning, community engagement, and personal branding is emphasized for career growth, especially as QE practices evolve in regulated industries like banking and government, where AI adoption remains cautious and incremental.
What If
-
What if you built an AI-augmented test case generator tailored to your niche product?
- Move: Identify a high-friction, repetitive testing area in your current project (e.g., form validation, API contract checks). Use ChatGPT or a local LLM to generate 10 - 20 test cases based on a precise prompt you craft (e.g., “Generate edge-case test inputs for a user registration API with phone number validation in Canada”). Integrate output into your test suite and refine manually.
- Why Now?: AI can cut test design time from days to minutes, and solo developers can leverage this to ship faster while maintaining coverage - especially as AI tooling becomes more accessible and prompt engineering skills offer immediate ROI.
- Expected Upside: Reduce manual test creation by 50 - 70% in targeted areas, improve edge-case coverage, and free up time to focus on exploratory testing or product improvements.
-
What if you implemented a trust-first checklist for any AI feature you build or test?
- Move: Create a lightweight, reusable checklist (e.g., in Markdown or Notion) with criteria like: “Does this AI output hallucinate?”, “Is bias checked against 2+ demographic inputs?”, “Is there a fallback for incorrect output?”, “Is user data encrypted in transit and at rest?”. Apply it to your next AI-integrated feature or prototype.
- Why Now?: With only 10% of AI pilots making it to production, early validation of trust factors (security, bias, reliability) prevents costly rework and builds user confidence - critical for solo operators with limited recovery bandwidth.
- Expected Upside: Avoid launching broken or unethical AI features, increase stakeholder trust, and differentiate your product in markets wary of AI risks (e.g., fintech, govtech).
-
What if you launched a public, AI-powered QA portfolio site to showcase your skills and attract opportunities?
- Move: Use AI tools like Lovable or ChatGPT + Vercel to generate a clean, responsive personal site in <48 hours. Include sections: projects, testing frameworks you’ve built, AI-assisted test examples, and speaking/writing links. Deploy it, add analytics, and share it on LinkedIn with a post about your QA philosophy.
- Why Now?: Personal branding compounds over time, and AI now enables solo developers to build professional sites in hours instead of weeks - just as demand for skilled QA professionals with AI fluency is rising.
- Expected Upside: Attract freelance clients, speaking invites, or job offers; establish authority in the QA community; create a living resume that demonstrates both technical and strategic skills.
Takeaway
- Integrate AI-powered test case generation into your workflow, but apply precise prompts and human validation to avoid irrelevant or excessive outputs.
- Focus on solving specific, high-impact business problems before adopting AI tools - validate success on a small scale before expanding.
- Build a personal portfolio website using AI tools like ChatGPT for content and design, but manually refine the output to ensure quality and professionalism.
- Prioritize risk-based testing strategies and deepen domain knowledge to stay relevant as AI automates repetitive QA tasks.
- Join active QA communities (e.g., Keyway) and pursue targeted certifications (e.g., MIT, Google) to stay updated on AI advancements and expand professional visibility.
For a PDF of longer Software Testing Podcast Episode Summaries with Briefing Notes and more detailed summary notes, visit EvilTester Patreon Podcast Summaries.