The STEMINISTS Podcast

Everyone loves to say RAG is the answer to AI hallucinations. But what if the real answer is a little more complicated? 

In this episode of The STEMINISTS Podcast, Phoebe Goh sits down with Ramya Ravi, AI Developer Advocate at NetApp Instaclustr, to unpack what Retrieval-Augmented Generation (RAG) actually does, why it's become a cornerstone of enterprise AI, and where organizations are still getting it wrong.  

Together, they demystify the mechanics behind RAG, from embeddings and vector search to semantic search and hybrid search. Ramya explains why RAG isn't magic (it's math), how it helps AI systems find relevant business knowledge, and why simply implementing RAG doesn't automatically eliminate hallucinations.  

The conversation also explores one of the biggest challenges facing AI projects today: data quality. From duplicate documents and stale content to weak evaluation practices, Ramya shares practical advice on what organizations should tackle before rushing into a RAG implementation.  

Looking ahead, Phoebe and Ramya discuss long context windows, the next evolution of enterprise AI, and why the future may not be a choice between RAG and large context models, but a powerful combination of both.  

In this episode: 

What RAG actually is and how it works behind the scenes 

Why RAG reduces hallucinations but doesn't eliminate them 

The difference between keyword, semantic, and hybrid search 

Why data quality is often the hidden factor in AI success 

How to properly evaluate a RAG implementation 

When to use long context windows versus retrieval systems 

Why the future of enterprise AI will likely rely on both approaches  

If you've been told that RAG is the silver bullet for enterprise AI, this episode offers a smarter, more practical perspective on what it can do, what it can't do, and how to make it work in the real world. 

 

Connect with us! 

https://www.linkedin.com/in/ramya-ravi19/

https://www.linkedin.com/in/mekkacodes/ 

https://www.linkedin.com/in/phoebegoh/

Fan Favorite Replay: As AI continues to reshape how organizations work, lead, and grow, we're bringing back one of our most popular conversations on preparing people and businesses for the future. The insights in this episode are just as relevant today as when it first aired, making it a must-listen for leaders navigating the age of AI. 

Are you ready to lead in a world transformed by AI? You'd better be, and this episode will help you get there. 

Hosts Phoebe Goh and Mekka Williams sit down with Sarah Bartel, Accenture's Global Cloud Talent and Organizational Change Lead, for an engaging discussion about what it takes to prepare teams, cultures, and organizations for the future of work. Drawing on her extensive experience helping organizations navigate technology-driven transformation, and her role co-teaching MIT's Leading the AI Driven Organization course, Sarah offers practical strategies for leading through change while keeping people at the center. 

Together, they explore how leaders can build more adaptable organizations, why traditional approaches to talent management are evolving, and what it takes to create a workforce ready to thrive alongside AI. You'll learn why skills-based hiring and team design are becoming critical competitive advantages, how trust serves as the foundation for successful transformation, and why AI works best when it's used to elevate human potential rather than replace it. 

Key Takeaway: The future belongs to organizations that invest in both technology and people. Leaders who embrace continuous learning, build cultures of trust, and focus on skills over job titles will be best positioned to succeed in an AI-powered world. 

Whether you're leading a large-scale transformation, building the next generation of talent, or simply looking to become a more effective leader in a rapidly changing environment, this fan-favorite episode is packed with actionable insights and fresh perspectives. 

Tune in and discover how to lead with confidence, adaptability, and a people-first mindset in the AI era. 

 

Work, Workforce, Workers Age of Generative AI Report 

https://www.accenture.com/content/dam/accenture/final/accenture-com/document-2/Accenture-Work-Can-Become-Era-Generative-AI.pdf  

 

Leading the AI Driven Organization MIT Course  

https://executive.mit.edu/course/leading-the-ai-driven-organization/a054v00000r9U5cAAE.html  

 

Develop your skills with NetApp  

https://www.netapp.com/support-and-training/netapp-learning-services/learning-paths/  

 

Connect with us! 

https://www.linkedin.com/in/sarah-gottry-bartel/  

https://www.linkedin.com/in/mekkacodes/ 

https://www.linkedin.com/in/phoebegoh/

DORA may be a financial services regulation, but don’t let the name fool you, it’s rapidly becoming everyone’s problem. 

In this episode of The STEMINISTS Podcast, Phoebe Goh sits down with Katy Rankin, Cybersecurity Counsel at NetApp, to unpack the Digital Operational Resilience Act (DORA) and why technology vendors are suddenly finding themselves at the center of some very detailed and very urgent customer conversations.  

From supply chain risk and subcontractor transparency to regulatory scrutiny and personal liability, Katy explains why the days of simply pointing to a compliance certificate and calling it a day are long gone. As financial institutions race to meet DORA requirements, they're asking tougher questions of their vendors and expecting far more specific answers. The catch? Those answers aren't always easy to find.  

Phoebe and Katy explore what DORA is really trying to achieve, why regulators are worried about a single cyber incident taking down multiple financial institutions at once, and how organizations can move beyond box-ticking exercises toward meaningful operational resilience. They also discuss the growing role of cybersecurity counsel, the challenges of managing third-party risk at scale, and why asking better questions is often the key to getting better answers.  

In this episode: 

Why DORA is much more than "just another compliance requirement" 

The rising importance of cybersecurity counsel in modern organizations 

How third-party and supply chain risk have become board-level concerns 

What financial institutions should be asking their technology providers 

The balance between transparency, security, and practical risk management 

Why operational resilience is ultimately about keeping critical services running when things go wrong  

And because no STEMINISTS episode is complete without an unexpected twist, Katy shares a fascinating archaeological discovery. History, it turns out, may have had its own STEMINISTS.  

As always, make sure to like, follow and share with your inner circle.  

 

Learn More:

https://www.netapp.com/industries/financial-services/dora/

 

Connect with us! 

https://www.linkedin.com/in/katyrankin/  

https://www.linkedin.com/in/mekkacodes/ 

https://www.linkedin.com/in/phoebegoh/

Fan Favorite Replay: Originally one of our most talked-about episodes, we're bringing this conversation back because the insights are even more relevant today. As AI continues to transform the threat landscape and security leaders face mounting pressure from every direction, this episode remains a must-listen for CISOs, security professionals, and business leaders alike. 

Being a CISO these days feels straight out of Mission Impossible... stopping the bad actors without the luxury of self-destructing instructions, fancy gadgets, or dramatic warning music. The stakes are sky-high, and the challenge of keeping an organization secure can often feel impossible. As AI rewrites the rules of business, security leaders find themselves navigating a perfect storm of shrinking budgets, expanding attack surfaces, and increasingly sophisticated threats. 

In this fan-favorite episode, we cut through the noise with Shira Rubinoff, CEO of The Cybersphere Group, for a candid and eye-opening look at what CISOs are really up against. From deepfakes that can fool even seasoned professionals to shadow AI tools being used across organizations without oversight, Shira shares practical insights on the realities of modern cybersecurity leadership. 

Shira breaks down the essential truths every security leader needs to understand: 

Bolting on security after the fact is a guaranteed failure. 

Your people can become your strongest line of defense through trust, awareness, and effective training. 

AI is not here to replace security professionals, but to amplify their capabilities and effectiveness. 

If you want to understand what really keeps CISOs up at night and how the best leaders are staying ahead of rapidly evolving threats, this is an episode you won't want to miss. 

 

Connect with us! 

https://www.linkedin.com/in/mekkacodes/ 

https://www.linkedin.com/in/phoebegoh/ 

https://www.linkedin.com/in/shirarubinoff/

AI Is Too Expensive. AI Is Free. AI Is Taking All the Jobs. Humans can’t be replaced. Which Is It? 

If your AI news feed is giving you whiplash, you're not alone. 

In this special AI Mythbusters episode of The STEMINISTS Podcast, hosts Phoebe Goh and Mekka Williams put some of the biggest AI headlines under the microscope and aren't afraid to call out where hype is masquerading as fact.  

First up: Is free AI really disappearing? Phoebe and Mekka unpack the reality behind "free" and "freemium" AI services, exploring the hidden costs of compute, the tradeoffs around privacy and data usage, and why the future may belong to smaller, smarter, and more efficient AI models not just the biggest ones. 

Then they tackle a myth that's dominating boardrooms everywhere: Is AI actually cheaper than human labor? From data center investments and infrastructure costs to training, governance, process redesign, and the ongoing need for human oversight, they reveal why comparing AI to people isn't nearly as simple as the headlines suggest.  

Along the way, they make the case for something that never goes out of style: good data practices. Because whether you're all-in on agents and automation or just experimenting with AI, success still starts with knowing your data, protecting it, and managing it well.  

Expect myth-busting, practical insights, a few laughs, and a refreshing reminder that in the race toward AI adoption, humans are still very much part of the equation.  

Topics covered: 

Why "free" AI isn't really free 

The real economics of AI compute and infrastructure 

Data privacy, security, and governance 

Human-in-the-loop AI and agentic workflows 

Small models vs. massive models 

Building a data strategy that survives the hype cycle 

 

And remember: if an AI headline sounds too dramatic, too optimistic, or too terrifying to be true... it probably deserves a little myth-busting of its own. As always, please remember to like, follow and share.  

 

Connect with us! 

https://www.linkedin.com/in/mekkacodes/ 

https://www.linkedin.com/in/phoebegoh/