How to Evaluate AI You Can Trust E166

How to Evaluate AI You Can Trust E166

Author: Punit Bhatia | Data Privacy, Sourcing & EU AI Act Specialist | ISO Standards June 4, 2026 Duration: 31:22

How do you know if an AI system is trustworthy, compliant, ethical, and fit for purpose? 
 
In this episode of the FIT4Privacy Podcast, Punit Bhatia is joined by Stella Liu, an AI evaluation expert and founder of AI Evals & Analytics, to unpack one of the most practical and overlooked challenges in AI today: how to evaluate AI systems before and after deployment. 

 
KEY MOMENTS 

02:09 —AI Definition 

03:02 —AI Evaluations 

10:31 — Why AI Testing Is Hard 

14:06 — Evals Plus Analytics 
18:15 —Synthetic Data 

23:47 —Protecting Privacy Ethical 
29:05 — AI Evals as a Company 

29:52 —How to reach Stella Liu 


Stella explains why AI behaves differently from traditional software, why testing code alone is no longer enough, and how AI evaluations (AI evals) help organizations assess real-world behavior, risk, and performance. From evaluation driven development to continuous monitoring in production, the conversation explores how teams can move beyond guesswork and hype toward repeatable, measurable AI governance. 

 
⸻ 

ABOUT THE GUEST 

Stella Liu is the Co-founder of AI Evals & Analytics (Maven), where she created the AI Evals & Analytics Playbook and teaches top-rated courses on LLM evaluation, monitoring, and product alignment. She’s also the Head of AI Applied Science at ASU, leading evals and analytics across university-wide AI products and building higher-ed’s first formal AI evaluation framework, and she previously led data science at Shopify and Carvana with 12+ years shipping large-scale ML systems. 

 

ABOUT THE HOST    

Punit Bhatia is one of the leading privacy experts who works independently and has worked with professionals in over 30 countries. Punit works with business and privacy leaders to create an organization culture with high privacy awareness and compliance as a business priority. Selectively, Punit is open to mentor and coach privacy professionals.  

 

⸻ 

Resources & Links 

 

Guest Links 
Stella Lui 

• Website: https://maven.com/  
• LinkedIn: https://www.linkedin.com/in/wenxingl/ 

 

Grow Skills (Privacy Courses & Insights) 

• Courses: https://growskills.store/courses/ 

• Insights: https://growskills.store/insights/ 

• Website: https://growskills.store/ 

 

FIT4Privacy 

• Website: https://www.fit4privacy.com 

• Podcast: https://www.fit4privacy.com/podcast 

• Blog: https://www.fit4privacy.com/blog 

• YouTube: http://youtube.com/fit4privacy 

 

Punit Bhatia 

• Website: https://www.punitbhatia.com 

 

Books  

• Be Ready for GDPR 

• AI & Privacy – How to Find Balance 

• Intro to GDPR 

• Be an Effective DPO 



Punit Bhatia, an expert in artificial intelligence, privacy, and sourcing, hosts The FIT4Privacy Podcast-AI and Privacy insights in collaboration with Grow Skills Store. Here, the often complex and rapidly changing worlds of data protection and AI are broken down into clear, practical conversations. Each episode delves into specific frameworks like GDPR and ISO standards, while also examining the broader ethical implications and the critical need for digital trust in modern business. The discussions go beyond theory, focusing on actionable insights that professionals can apply directly to their own management and operational challenges. This podcast serves as a vital resource for business leaders and managers navigating this landscape, featuring regular conversations with a variety of industry experts. The dialogue is grounded in real-world application, whether discussing evolving EU regulations or the practical intersection of AI and privacy law. Recognised for its quality, the series consistently ranks highly in its categories. For anyone looking to deepen their understanding of these crucial topics, this podcast offers a reliable and insightful guide through the complexities of compliance, ethics, and technology.
Author: Language: en-gb Episodes: 50

The FIT4Privacy Podcast | Exploring ISO Standards AI Privacy | Grow Skills Store
Podcast Episodes
Will all developer lose job because of AI doing the coding [not-audio_url] [/not-audio_url]

Duration: 7:57
Will AI replace developers or redefine what it means to be one?In this episode, Punit and Jake explore one of the most debated questions on tech today. While AI can now generate code, automate repetitive tasks, and build…
Human First AI Second : A Practical Approach to Success E168 [not-audio_url] [/not-audio_url]

Duration: 26:26
What does it mean to stay human in an AI‑first world? Is there any news today without AI? As AI dominates media and journalism, what does that mean for trust, attention, and how news is produced? AI is everywhere in toda…
How AI Is Reshaping Media and Journalism [not-audio_url] [/not-audio_url]

Duration: 8:01
How is AI transforming media and journalism and what does it mean for trust and authenticity? In this episode, Punit and Delia explore how the media landscape has evolved from traditional journalism to a digital, click-d…
How Tech is Invading Privacy E167 [not-audio_url] [/not-audio_url]

Duration: 25:53
Technology is evolving rapidly and with it, the boundaries of personal privacy are being tested. Privacy, security, and user choice are the foundations of responsible technology. When AI and cameras are everywhere, where…
AI: Power, Threats & Responsibility E172 [not-audio_url] [/not-audio_url]

Duration: 22:45
What does responsible AI look like in practice? In this episode, we explore AI governance, AI risks, ethical AI, and how businesses can use artificial intelligence responsibly to create value without losing trust. KEY MO…
AI is Just Getting Started [not-audio_url] [/not-audio_url]

Duration: 7:01
AI is everywhere today but what if we’re only at the very beginning? In this episode, Punit and Oscar explore why current AI tools are just the entry level of what’s to come, and how the next decade could transform every…
AI Evaluation Methods [not-audio_url] [/not-audio_url]

Duration: 7:07
How do you know if your AI is truly trustworthy, compliant, and ready for real-world use?In this episode trailer, we explore the evolving field of AI evaluation from understanding why AI systems can behave unpredictably…
AI on LinkedIn Authenticity, Automation & the New Rules E165 [not-audio_url] [/not-audio_url]

Duration: 29:19
AI is changing how professionals communicate, market, and build relationships on LinkedIn but using more automation doesn’t always lead to better outcomes. In this episode of the FIT4Privacy Podcast, we sit down with Jud…
AI on LinkedIn Authenticity Beats Automation [not-audio_url] [/not-audio_url]

Duration: 8:44
Is AI making LinkedIn outreach more effective or less trustworthy?In this trailer, we explore how AI is reshaping marketing, outreach, and personal branding on LinkedIn and why authenticity matters more than ever.You’ll…
How CalPrivacy is Leading Privacy in the US E164 [not-audio_url] [/not-audio_url]

Duration: 41:23
In this episode of the FIT4Privacy Podcast, we speak with Tom Kemp, Executive Director of the California Privacy Protection Agency (CalPrivacy), about why California continues to set the direction for privacy regulation…