The STEMINISTS Podcast

If AI is only as good as the data behind it, we've got a problem. 
In this episode of The STEMINISTS Podcast, host Phoebe Goh welcomes back Darnell Fatigati to talk about one of the least glamorous but most important topics in AI: dark data. You know, the forgotten files, mystery folders, archived documents, and digital clutter quietly piling up across organizations while everyone races to build the next AI breakthrough. 
The catch? AI doesn't know the difference between your best data and your messiest data. 
Darnell breaks down why dark data has become one of the biggest barriers to successful AI adoption, how it can introduce security, compliance, and reliability risks, and why companies need to stop treating data governance like a future problem. From AI copilots to autonomous agents, the smarter our systems get, the more dangerous unmanaged data becomes. 
The conversation explores what it takes to uncover hidden data, build trust in the information that fuels AI, and create a foundation that can actually support AI at scale, without creating new headaches along the way. 
And because STEMINISTS never leaves you without a fun fact for your next team meeting, stick around for a fascinating look at how AI helped researchers discover more than a thousand previously overlooked objects hiding in decades of astronomical data. 
Key Takeaways: 
Your biggest AI risk might be data you didn't even know you had. 
Bad data governance doesn't just create compliance problems. It creates bad AI outcomes. 
Before chasing the next AI trend, make sure your data isn't living in a digital junk drawer. 
Like what you heard? Share this episode with the colleague who's always talking about AI but never talks about data. Follow The STEMINISTS Podcast for more smart conversations on technology, innovation, and the people shaping what's next. And leave a comment telling us what topic we should unpack next. 
 
Connect with us! 
https://www.linkedin.com/in/darnellfatigati/  
https://www.linkedin.com/in/phoebegoh/ 
 
-----
 
Phoebe Goh and Darnell Fatigati on dark data: the enterprise's hidden AI liability 
Episode overview 
In this episode of "The Steminists," host Phoebe Goh talks with Darnell Fatigati, principal product marketing manager at NetApp, about "dark data," information an enterprise collects, stores, and protects but never actively uses, classifies, or governs. Darnell Fatigati describes dark data as the enterprise's data estate blind spot: archived emails, file shares, logs, backups, old folders, collaboration tools, images, and videos that nobody can answer basic questions about, including who owns the data, whether it's sensitive, and whether AI can use it safely. 
The central argument from Darnell Fatigati is that dark data becomes dangerous the moment AI tries to use it. Unmanaged data forces an enterprise to determine accuracy, sensitivity, access rights, and context after the fact, and that discovery work carries real cost. Darnell Fatigati distinguishes the stakes by AI type: generative AI built on dark data produces weak answers and hallucinations, while agentic AI built on dark data can trigger harmful actions across business systems, since agents plan, reason, call tools, and act. 
Darnell Fatigati proposes a practical fix: start small, discover and classify data at the storage infrastructure level, and treat the work as continuous rather than a one-time project. Darnell Fatigati frames data management, not the AI model, as the hardest part of moving enterprise AI from pilot to production, and closes the conversation with an analogy comparing dark data to "jellyfish galaxies" recently discovered by AI in decades-old Hubble Space Telescope images. 
Key takeaways 
The following takeaways summarize the practical guidance Darnell Fatigati shared for enterprises managing AI and data infrastructure. 
Dark data is information an enterprise collects, stores, and pays to protect, but doesn't actively use, classify, govern, or learn from. 
Dark data commonly includes archived emails, file shares, logs, backups, old folders, collaboration tools, images, and video documents. 
Dark data appears mostly in unstructured data sources, where metadata exists but remains undiscovered and unorganized. 
When AI activates unmanaged data, the enterprise must determine its accuracy, sensitivity, ownership, and context, and that discovery work carries real cost. 
With generative AI, poor data quality shows up as weak answers, hallucinations, and inconsistent summaries. 
With agentic AI, poor data quality creates operational risk because agents plan, reason, call tools, and trigger actions across business systems. 
Dark data risks include security leaks of sensitive information, data privacy violations, compliance failures tied to retention and access policies, rising costs, and degraded AI outputs. 
Ownership of dark data is fragmented across an organization, which is why Darnell Fatigati says it becomes "everybody's problem" and therefore "nobody's priority." 
The recommended starting point is discovery and classification at the storage infrastructure level, since that's where data lives and changes. 
Managing dark data should be continuous, not a one-time project, because data moves, gets accessed, and changes constantly. 
Solving dark data requires collaboration between storage, security, and AI teams, with each team contributing what it needs to know about the data. 
The hardest part of enterprise AI isn't the model. It's giving AI trusted, governed, high-quality access to the right data at the right time. 
Moving from AI pilot to production requires AI to operate on messy, distributed, constantly changing enterprise data instead of small, curated pilot datasets. 
Topics and themes 
This episode covers the risks dark data poses to enterprise AI and the practical steps organizations can take to discover, classify, and govern it. 
Dark data definition: unused, unclassified, and ungoverned enterprise information sitting across structured and unstructured sources. 
AI activation risk: how pointing AI at unmanaged data forces sudden, costly decisions about accuracy, sensitivity, and access. 
Generative AI versus agentic AI stakes: weak answers and hallucinations versus operational actions that can harm customers or business systems. 
Security, privacy, and compliance exposure: sensitive or regulated information sitting unmonitored until an AI agent accesses it. 
Fragmented ownership: why dark data responsibility spans storage, security, and AI teams without one clear owner. 
Infrastructure-level discovery: classifying and enriching metadata at the storage layer as a continuous, team-based process. 
Pilot-to-production scaling: the shift from curated, manually supported pilot data to messy, distributed, constantly changing production data. 

Everyone's racing to deploy AI agents. Few are asking what happens when those agents have access to your most valuable data. 
On this episode of The STEMINISTS Podcast, hosts Phoebe Goh and Darnell Fatigati sit down with Sandra Dunbar, NetApp's lead for cyber resilience product marketing, to talk about the security challenge hiding behind the AI hype. As organizations give AI agents more autonomy, the real risk isn't a rogue machine plotting world domination. It's a helpful agent moving at machine speed with access it should never have had in the first place. 
From zero trust and identity management to data governance and cyber resilience, Sandra explains why the future of AI security isn't about trusting agents more. It's about trusting them less, intelligently. The conversation explores how organizations can embrace agentic AI without sacrificing control, visibility, or common sense. 
And yes, there's a surprisingly accurate comparison between AI agents and teenagers. Both need boundaries. Both think they know what they're doing. And both can create a lot of trouble when left unsupervised. 
 
Key Takeaways: 
✅ The biggest AI risk isn't malicious intent. It's excessive access. Well-meaning AI agents can create outsized problems when they're given broad permissions and unclear guardrails. 
✅ Zero trust isn't just for humans anymore. AI agents need identities, permissions, monitoring, and accountability just like any employee or application. 
✅ Cyber resilience is the ultimate safety net. Even with strong controls in place, immutable data protection, recovery capabilities, and layered defenses remain essential when AI-driven mistakes happen. 
Enjoyed the conversation? Share this episode with a colleague, teammate, or fellow tech enthusiast who's navigating the opportunities and risks of AI. If you found value in the discussion, be sure to follow The STEMINISTS Podcast so you never miss an episode. 
And we'd love to hear from you: What topics, trends, or guests should we cover next? Drop a comment, leave a review, or connect with us on social. Your ideas help shape future conversations. 
Because the best discussions don't end when the episode does. They start there. 🎙️✨ 
 
Connect with us!
https://www.linkedin.com/in/darnellfatigati/  
https://www.linkedin.com/in/phoebegoh/ 
 
-----
The STEMINISTS: Securing AI agents with Sandra Dunbar of NetApp 
Episode overview 
In this episode of "The STEMINISTS," hosts Phoebe Goh and Darnell Fatigati interview Sandra Dunbar, NetApp's lead for cyber resilience product marketing, about securing AI agents. Sandra Dunbar specializes in cyber resilience, security, and agentic AI [1:02]. The conversation covers the balance between agent autonomy and security guardrails, real-world risks, implementation challenges tied to the Model Context Protocol, and best practices for deployment, including zero trust principles, access controls, and monitoring. 
Sandra Dunbar's central argument: well-intentioned AI agents create more risk than malicious actors. An agent with excessive access can execute thousands of actions in seconds without human judgment, and it can repeat the same mistake across thousands of files, systems, and transactions at machine speed [1:33] [3:51] [4:08]. Sandra Dunbar illustrates this risk with the PocketOS incident, in which an agent lacking proper trust boundaries deleted a production database and its backup while trying to complete a task, not while trying to cause damage [2:54] [3:15]. 
Sandra Dunbar also outlines operational safeguards for enterprise AI deployment: zero trust verification of agent identity, restricted access through role-based access control (RBAC) and attribute-based access control (ABAC), human approval workflows for high-risk actions, behavioral monitoring, and immutable, isolated data copies for recovery [6:48] [7:04] [7:56] [8:17] [8:40] [9:43] [10:12]. Near the end of the interview, Sandra Dunbar draws a parenting analogy, comparing guardrails for her newly independent daughter to guardrails for AI agents, a comparison that became the episode's most memorable moment [12:00] [12:20]. 
Key takeaways 
These takeaways summarize Sandra Dunbar's core guidance on securing AI agents in enterprise environments. 
The greatest security risk with agentic AI comes from well-intentioned agents that have too much access, not from malicious actors [1:33]. 
Agents can repeat the same error across thousands of files, systems, and transactions in seconds, while a human typically makes one mistake at a time [3:51] [4:08]. 
The PocketOS incident shows what happens without proper trust boundaries: an agent deleted a production database and its backup while attempting to complete an assigned task [2:54] [3:15]. 
Microsoft CEO Satya Nadella's security advice, as paraphrased by Sandra Dunbar, is to assume every model and every agent is compromised and to build in an emergency stop beforehand [1:50]. 
The Model Context Protocol (MCP) standardizes how agents connect to tools and data, but that standardization creates a high-value trust boundary that organizations must secure explicitly [5:53] [6:22]. 
Zero trust principles apply directly to AI agents and require even stricter identity verification and monitoring than organizations typically apply to human users [6:48] [7:04]. 
Effective agent security requires a layered approach: RBAC, ABAC, identity management, data-layer access controls, behavioral analytics, and immutable backups [7:56] [8:17] [8:40] [9:43]. 
Human oversight remains essential through approval workflows and administrator alerts for sensitive actions, since agents themselves can't receive multi-factor authentication prompts directly [10:12] [10:34] [10:42]. 
Immutable, isolated data copies give organizations a recovery path when an agent makes an error [8:40]. 
Sandra Dunbar's parenting analogy frames multiple interacting agents as "friends" with collectively worse judgment, a model for understanding multi-agent risk [12:20]. 
Topics and themes 
The episode centers on four recurring themes: balancing agent autonomy with control, machine-speed risk, zero trust architecture, and layered security design. 
Balance between agent autonomy and security control, discussed throughout the interview [0:27] [0:40] [2:23] [10:12]. 
Machine speed versus human judgment as a distinct risk category for AI agents compared to human error [4:08] [4:47]. 
Zero trust architecture applied specifically to non-human identities, including agents [6:40] [6:48] [7:04] [9:43]. 
Layered security ecosystem design, combining access controls, identity management, and monitoring [9:11] [9:26] [9:43] [11:05]. 
The Model Context Protocol (MCP) as infrastructure that requires its own security boundary [5:53] [6:22]. 
Agent-specific behavioral analytics, extending user behavior analytics originally built for human users [8:57] [9:11] [9:26]. 
Parenting and child-rearing analogies used to make agent governance concepts accessible to non-technical audiences [12:00] [12:20] [12:40] [13:02] [13:15]. 
Notable quotes 
"The greatest risk usually isn't a malicious agent. It's a well-intentioned agent that has too much access."—Sandra Dunbar [1:33] 
"You should assume every model, every agent is compromised and build in that emergency break beforehand."—Sandra Dunbar [1:50] 
"It wasn't trying to destroy data, it was trying to complete the task, but it had more authority than it had judgment."—Sandra Dunbar [3:15] 
"A human typically makes one mistake at a time... the agent can repeat the same mistake across thousands of files and systems and transactions in seconds."—Sandra Dunbar [3:51] [4:08] 
"MCP gives agents a doorway into your tools and your data, and the security challenge is making sure that that doorway doesn't become a shortcut to your existing control."—Sandra Dunbar [5:53] 
"If her friends come over, they're like additional agents coming into the house and they have very poor judgment. The more of them they get together, the worse the judgment gets."—Sandra Dunbar [12:20] 

Here's a hot take: AI isn't your biggest competitive advantage. Getting value from your technology might be. 
On this episode of The STEMINISTS, Phoebe Goh sits down with Deboja Das, VP of Customer Success Portfolio Management at NetApp, to talk about a role that's gone from behind-the-scenes support function to front-and-center business driver. 
As organizations invest heavily in AI, cloud, and cyber resilience, the pressure is on to prove those investments are actually moving the needle. Deboja explains why customer success is increasingly responsible for bridging the gap between ambitious technology strategies and tangible business outcomes. 
The conversation covers everything from proactive support and cyber preparedness to data governance and AI-driven customer experiences. The message is clear: technology may start the journey, but customer success determines whether you reach the destination. 
Key takeaways: 
Customer success is evolving from support center to strategic advisor. 
Data governance and security are more interconnected than ever. 
AI can help predict problems before they impact the business. 
Shared accountability drives better outcomes. 
Long-term value is the metric that matters most. 
Because in today's technology landscape, success isn't measured by what you buy. It's measured by what you achieve. 
 
Learn more: 
https://www.netapp.com/customer-experience/  
 
Connect with us!
https://www.linkedin.com/in/darnellfatigati/  
https://www.linkedin.com/in/phoebegoh/ 
 
---
Deboja Das on customer success in the AI era: why shared responsibility beats the support-ticket model 
Episode overview 
In this episode of "The STEMINISTS Podcast," host Phoebe Goh talks with Deboja Das, VP Customer Success Portfolio Management at NetApp, about how the relationship between vendors and customers has changed over the last ten to fifteen years. Deboja Das traces three distinct eras: reactive break-fix support, the SaaS and cloud era where customer success and product support worked closely on onboarding and adoption, and the current era, where professional services, customer success, product support, and digital capabilities combine to detect and resolve issues before customers even notice them. 
 
Deboja Das argues that data has become a critical business asset and, in many cases, the last line of defense for enterprises, which raises the stakes for how vendors help customers spot risks and blind spots. She points to infrastructure as a clear example: after years of attention shifting to the application layer, AI has forced a reinvestment in infrastructure because enterprise data has to become AI-ready. She also raises data sovereignty as a newly urgent concern, driven by geopolitical change, alongside long-standing questions of access, governance, and compliance. 
 
A second argument centers on shared responsibility for security and reliability. Deboja Das references the software update that disrupted the airline industry as evidence that failures stem from interconnected systems, not a single vendor's product issue. She concludes that security and uptime are a collective responsibility between vendor and customer, not a burden that falls on one side alone. She also describes specific NetApp practices, including success plans, intake forms, and an always-on value tracker and health dashboard, that put this philosophy into operation. 
 
Key takeaways 
Deboja Das offered guidance for technology buyers and vendors navigating the shift from reactive support to proactive, outcome-driven partnerships. 
Customer success has moved from reactive problem-fixing to measuring "time to value," meaning vendors now focus on how fast a customer's business problem gets solved rather than how fast a support ticket gets closed. 
Infrastructure has returned as a priority investment area because AI requires enterprise data to be reorganized and prepared as AI-ready data. 
Data sovereignty has become a bigger issue recently due to geopolitical changes, even though it wasn't a major topic a few years ago. 
Security and system reliability are a shared responsibility between vendor and customer, not the sole obligation of either party. 
Modern customer success integrates professional services, product support, and digital capabilities into one team, aiming to detect and resolve issues before the customer notices them. 
NetApp uses success plans, intake forms, and an always-on value tracker and health dashboard to show customers ongoing value, even when nothing has broken. 
The single most useful question a customer can ask a vendor is how the vendor will keep the customer successful on an ongoing basis, not just at the point of sale or renewal. 
 
Topics and themes 
This episode covers the shift in vendor-customer relationships toward shared accountability for data, security, and business outcomes. 
The evolution of customer success: from break-fix support, through SaaS-era onboarding, to today's integrated, proactive model. 
Data as a critical business asset: the growing responsibility to secure, govern, and track who accesses enterprise data. 
AI-ready infrastructure: the resurgence of infrastructure investment driven by AI's data requirements. 
Data sovereignty and compliance: an emerging concern shaped by geopolitical shifts. 
Shared responsibility: the argument that vendors and customers jointly own security and reliability outcomes, illustrated by a major software-caused airline industry outage. 
Outcome ownership: measuring success by whether a customer's original business problem got solved, not by response speed. 
Always-on value demonstration: tools like health dashboards and value trackers that show benefit continuously, not only during a crisis. 
Vendor evaluation: the questions buyers should ask to assess a vendor's commitment to the full customer life cycle. 
 
Notable quotes 
"I'm still an engineer at heart, even though I'm now talking business and top line and margins, but I'm still an engineer. And that's what makes this role very interesting because the promise of engineering is what is delivered by customer success." (0:54) 
"Previously it's like, 'Hey, I bought a product, it's not working, can you fix it?' ... Right now it's almost like we don't want to talk about issues. We don't want to talk about how fast we are resolving your problem. We are talking about the outcome, first time to value." (7:46) 
"That's always been a question. Securing a data seems like to be a customer's responsibility." (5:00) 
"I think it's a collective responsibility of everybody. It is as much a customer's responsibility as it's a vendor's responsibility." (5:48) 
"We go that extra mile to make sure that full life cycle capability, we got your back for you." (9:39) 
"How are you ensuring that I'm successful on an ongoing basis, not just the first time? I sell you the product and then I'm gone and I just come back to you on  

The AI conversation has changed. The hype is loud. The hard questions are louder. 
Can your infrastructure keep up? Can your data be trusted? And are you building for long-term value or just chasing the latest shiny object? 
In this episode of The STEMINISTS Podcast, Phoebe Goh, joined by guest co-host Darnell Fatigati and technology strategist Kris Cornwall, unpack the biggest takeaways from NetApp Insight 2026 and the technology shifts shaping the next chapter of enterprise AI. 
Together, they explore the debut of NetApp Novus, the rise of AI factories, the growing demand for flexibility across storage environments, and why interoperability may be one of the most underrated competitive advantages in tech today. 
If your organization is trying to move from AI curiosity to AI capability, this conversation offers a refreshingly practical perspective, with just enough healthy skepticism to keep things interesting. Because while everyone is talking about AI, the organizations creating real value are focused on something far less glamorous: getting their data house in order. 
Key Takeaways 
AI success starts with trusted data. The biggest barrier to AI isn't the model. It's knowing where your data lives, whether it's governed, and if it's ready for production. 
The future isn't homogeneous. Enterprises need the flexibility to leverage data across clouds, on-premises environments, and multiple storage platforms without costly disruption. 
Platform strategy beats point solutions. As AI, cloud, virtualization, and security continue to converge, organizations gain more value from interoperable ecosystems than isolated technologies. 
Tune in to hear why the next wave of AI leadership won't be defined by who adopts AI first, but by who builds the strongest data foundation beneath it. From infrastructure strategy to practical execution, this episode offers a candid look at what it really takes to turn AI potential into business impact. 
 
Learn More: 
https://www.netapp.com/novus/ 
https://www.netapp.com/insight/  
https://www.netapp.com/platform/  
 
Connect with us! 
https://www.linkedin.com/in/darnellfatigati/  
https://www.linkedin.com/in/phoebegoh/ 
 
The STEMINISTS Podcast: NetApp Insight 2026 recap: data readiness, platform strategy, and the next phase of enterprise AI 
Episode overview 
In this episode of "The STEMINISTS podcast," host Phoebe Go is joined by guest co-host Darnell Fatigati, and guest Kris Cornwall, a NetApp industry veteran. The conversation recaps NetApp Insight 2026, NetApp's annual in-person customer conference, and centers on the announcements shaping enterprise AI infrastructure. Kris Cornwall frames the event around a single question customers kept asking: not what AI can do, but how to operate it. 
The main argument from Kris Cornwall is that NetApp Insight 2026 announcements fit together under one umbrella, the NetApp platform, rather than standing as a collection of separate technologies. Kris Cornwall describes the platform as a way to unify storage, cloud, operations, security, and consumption into a single operating model, built to work with the technology choices customers have already made, including partnerships with Oracle, SAP, NVIDIA, Supermicro, Commvault, Nutanix, and Cisco. 
Darnell Fatigati brings the customer perspective into the discussion, pointing to data readiness as the issue infrastructure teams raised most often at the conference. Specific examples discussed include NetApp Novus, a storage architecture for gigawatt-scale AI factories and NeoCloud environments; support for heterogeneous storage that lets customers access data across formats and locations without copying it; a new partnership with Oracle Cloud Infrastructure as a fourth supported hyperscaler; and expanded virtualization support, including new integration with Nutanix alongside existing VMware and Red Hat options. 
Key takeaways 
The following takeaways summarize the main points Kris Cornwall and Darnell Fatigati raised about AI infrastructure readiness and the NetApp platform strategy. 
Customer conversations about AI shifted in 2026 from "what can AI do" to "how do I operate, secure, and fund it," according to both Kris Cornwall and Darnell Fatigati. 
Infrastructure and storage teams, not just data science teams, are now directly involved in AI planning conversations. 
Data readiness, not future scale, is the immediate barrier most enterprises face when deploying AI agents today. 
NetApp now supports heterogeneous storage, which means enterprises can connect SMB, S3, and NFS data across cloud and on-premises environments without migrating or copying it first. 
Avoiding data migration during AI integration reduces two specific risks: permissions not transferring correctly, and AI agents acting on outdated data. 
NetApp Novus is a storage architecture built for AI factories and gigawatt-scale NeoCloud environments running tens to hundreds of thousands of GPUs. 
NetApp added Oracle Cloud Infrastructure as a fourth supported hyperscaler, joining AWS, Google Cloud, and Microsoft Azure. 
NetApp® ONTAP® gives customers one consistent data foundation across hyperscalers, so a hybrid multicloud setup doesn't require four separate operating models. 
NetApp expanded virtualization support with a new Nutanix integration, in addition to existing support for VMware and migration paths to Red Hat. 
Kris Cornwall recommends that customers start AI planning with their workloads and desired outcomes, not with a specific product or announcement. 
Topics and themes 
This episode covers the shift from AI experimentation to AI operations, organized around the announcements made at NetApp Insight 2026. 
AI maturity curve: the move from exploratory AI questions in prior years to operational questions about security, funding, and integration in 2026. 
Data readiness: preparing existing data estates, across formats and locations, for use by AI agents without requiring migration. 
The NetApp platform: a unified operating model spanning storage, cloud, operations, security, and infrastructure consumption. 
Ecosystem partnerships: integrations announced with Oracle, SAP, NVIDIA, Supermicro, Commvault, Nutanix, and Cisco. 
NetApp Novus: a storage architecture built for gigawatt-scale AI factories and NeoCloud GPU environments. 
Heterogeneous storage support: in-place access to SMB, S3, and NFS data without copying or ETL processes. 
Multi-hyperscaler strategy: support for AWS, Google Cloud, Microsoft Azure, and now Oracle Cloud Infrastructure through NetApp ONTAP. 
Virtualization flexibility: support for VMware, Red Hat migration paths, and new Nutanix integration. 
Customer-led planning: starting AI initiatives from business workloads and outcomes rather than from specific products. 
Notable quotes 
"It was really exciting. To no surprise, AI was everywhere. That's what people are talking about. But also I feel like conversations were more grounded than they were a year ago." - Kris Cornwall, 00:01:28 
"Last year I felt like people didn't really know what questions to ask, but this year, definitely the storage, the infrastructure teams were leaning in saying, 'Okay, what does this mean for me from an infrastructure standpoint?'" - Darnell Fatigati, 00:01:44 
"We're not saying you need to move this data, so you don't have to copy it." - Darnell Fatigati, 00:07:00 
"We introduced NetApp Novus, a new storage architecture that is really geared for the highest, we call them AI factories or gigawatt scale environments." - Kris Cornwall, 00:05:05 
"We know that most enterprises don't use one hyperscaler and they make lots of workload decisions and about where to place their data." - Kris Cornwall, 00:09:00 
"I think customers need to start with their workloads and the outcomes they want to get. Don't start with the product, don't start with an announcement, but start with their workloads." - Kris Cornwall, 00:10:47 

Most infrastructure conversations start with budgets. The best ones start with outcomes. 
If you've ever thought Storage as a Service was just a finance conversation, this STEMINISTS podcast episode might just change your mind. Or at the very least, make storage a lot more interesting. Phoebe Goh and Victoria Lam, Solutions Architect at NetApp, unpack why Storage as a Service should be viewed as a strategic enabler rather than simply a consumption model. From cloud transformation to AI readiness, Victoria explains why the organizations getting the most value are the ones planning for change, not just managing costs.  
Drawing on her experience helping global organizations, Victoria breaks down why forward-thinking leaders are looking beyond CapEx versus OpEx debates and focusing instead on agility, risk reduction, and long-term business outcomes. From unexpected AI growth and evolving workloads to hybrid cloud ambitions that never seem to arrive on schedule, she explains how Storage as a Service can help organizations stay ready without overcommitting to yesterday's plans.  
The conversation explores why so many companies still overbuy infrastructure, how consumption-based models can keep future cloud options open, and why successful technology strategies start with outcomes rather than hardware specifications. Because the smartest infrastructure decision isn't always, "What storage should I buy?" It's, "What problem am I actually trying to solve?"  
Plus, Victoria shares how her business background helps her translate between technical and business teams, offers insights from one of the world's most diverse technology regions, and reveals the unexpected truth that learning Australian slang might be more complicated than learning AI.  
If you've ever thought Storage as a Service was just a finance conversation, this episode might just change your mind. Or at the very least, make storage a lot more interesting. 🎙️ 
 
Key Takeaways 
Storage as a Service shifts infrastructure from a capital planning exercise to a strategic flexibility play. Instead of making long-term bets on future demand, organizations can adapt as business priorities evolve.  
The real ROI isn't just lower costs, it's lower risk. Consumption-based models help organizations avoid overprovisioning, reduce technology lock-in, and respond faster to changing business requirements.  
AI and data growth are making agility a board-level concern. As workloads become less predictable, infrastructure that can scale with demand becomes a competitive advantage.  
Cloud transformation requires optionality. For many organizations, Storage as a Service provides a practical bridge to hybrid cloud, keeping future migration paths open without forcing immediate decisions.  
Leading organizations start with outcomes, not infrastructure. The most effective technology strategies are built around business objectives such as growth, resiliency, and innovation, then aligned to the right operating model. 
If you enjoyed this episode, please follow, like and share.  
 
Learn More: 
https://www.netapp.com/novus/  
https://www.netapp.com/keystone/  
 
Connect with us!
https://www.linkedin.com/in/phoebegoh/ 

Everyone's obsessed with getting "good data." But here's the uncomfortable question: what happens when your AI has accurate data and still makes the wrong decision? 
In this episode of The STEMINISTS podcast, host Phoebe Goh welcomes back AI data expert Darnell Fatigati for a conversation that challenges one of the biggest assumptions in AI today: that data quality is enough. Spoiler alert: it ain’t.   
As AI agents move beyond answering questions and start making recommendations, triggering workflows, and taking action, the stakes get a lot higher. A data point can be technically correct and still completely miss the bigger picture. Without critical context, AI isn't making informed decisions. It's making confident guesses. And these guesses turn into mistakes that scale fast. 
Phoebe and Darnell unpack why data context, trust, governance, and even organizational memory are becoming essential ingredients for successful AI. They explore how businesses can help AI understand not just what happened, but why it happened, so agents can act responsibly instead of accidentally creating expensive chaos at machine speed.  
Key Takeaways: 
Why clean data doesn't automatically lead to smart AI decisions  
The difference between data quality and data integrity, and why it matters more than ever  
How context turns information into understanding and understanding into action  
Why organizational memory may be your next competitive advantage in AI  
Practical steps to make your data more AI-ready without boiling the ocean  
Plus, Darnell shares the story of an AI that managed a professional baseball game surprisingly well... until one very human variable showed up and reminded everyone that context still matters. ⚾🤖  
Because in the age of agentic AI, it's not enough for your data to be correct. Your AI needs to understand the assignment.  
Like what you hear? Follow and share The STEMINISTS Podcast with your network. The future of AI won't be built on more data alone. It'll be built on better understanding. 🚀 
 
Learn more about the Symbiotic relationship between AI and Data:
https://www.netapp.com/blog/symbiotic-relationship-data-and-ai/  
 
Check out how AI is only as good as the data that fuels it: 
https://www.netapp.com/blog/ai-is-only-as-good-as-the-data-that-fuels-it/  
 
Connect with us! 
https://www.linkedin.com/in/phoebegoh/ 
https://www.linkedin.com/in/darnellfatigati/  
 
-----
Let's Solve IT!: Darnell Fatigati on why agentic AI needs more than clean data 
 
Episode overview 
In this episode of "Let's Solve IT!," host Phoebe Goh talks with returning guest Darnell Fatigati about what changes when AI stops answering questions and starts taking action [0:00] [0:24]. Fatigati's central argument: agentic AI raises the bar for data, because systems that recommend decisions and execute tasks need more than accurate information. They need context [1:07] [1:23]. 
Fatigati draws a sharp line between data quality, whether information is correct, complete, consistent, and current, and data integrity, whether that information can be trusted in the moment it's used [2:07] [2:18]. A revenue figure can be completely accurate and still mislead an AI agent if it doesn't account for a sold business unit, a discontinued product, or a mid-quarter pricing change [2:18]. Without that context, Fatigati argues, very different business situations can look identical to an AI system [3:01] [3:14]. 
The conversation moves from definition to consequence. Fatigati explains that agentic AI doesn't just produce poor reports when data is flawed, it takes poor actions, and those actions can scale and propagate before anyone notices [4:05] [4:24]. To ground the argument, Fatigati cites a real-world example: an independent minor league baseball team that handed a full game to an AI manager. The system got almost every call right, until the starting catcher fell ill and the AI lacked the one current fact that mattered most [11:08] [11:27] [11:31] [11:54]. 
 
Key takeaways 
The following points capture the practical guidance Fatigati offered for organizations preparing data for agentic AI. 
Agentic AI raises the standard for data because systems are increasingly asked to recommend decisions and take actions, not just retrieve or summarize information [1:07] [1:23]. 
Data quality (correctness, completeness, consistency, currency) is necessary but not sufficient; data integrity, whether information can be trusted in context, is what determines good outcomes [2:07] [2:18]. 
Context transforms raw information into understanding, and its absence can make distinct situations look the same to an AI system [3:01] [3:14]. 
Agentic AI scales the consequences of poor data, since mistakes can propagate across business processes before a human intervenes [4:05] [4:24]. 
Agents need comprehension, not just access: they must know which sources are authoritative, how current the information is, what permissions apply, and when to escalate to a person [6:28] [6:45]. 
Giving an agent access to every company document doesn't create expertise, the same way giving a new employee access to every shared drive doesn't create expertise [6:28]. 
Organizations tend to preserve what decisions were made but lose the reasoning behind them, and that reasoning is increasingly valuable for training future AI systems [7:37] [7:48]. 
The recommended starting point for AI readiness is understanding existing data, not collecting more, beginning with a small number of high-value use cases [8:48] [9:40]. 
Knowledge graphs can help organizations connect distributed enterprise information and identify authoritative sources [9:16]. 
When organizations get data context right, AI moves past isolated productivity gains and becomes a participant in driving business outcomes [10:13]. 
Topics and themes 
This episode centers on what enterprise data needs to look like as AI shifts from answering questions to taking action. 
Context as the missing layer in enterprise AI: what data means, where it came from, how current it is, and what policies govern it [1:23] [2:18] [3:01] [3:14] [5:29] [6:45]. 
The gap between data access and data comprehension, for AI agents and for the organizations deploying them [6:28] [8:48] [9:00]. 
Trust as a prerequisite for autonomous action, grounded in visibility into agent reasoning, permissions, and controls [5:29]. 
Agentic AI as a distinct risk category, where repeated, scalable actions change the consequences of bad data [4:05] [4:24]. 
Organizational memory as an emerging priority: capturing the trade-offs and assumptions behind past decisions, not just the outcomes [7:37] [7:48]. 
Knowledge graphs as a mechanism for connecting enterprise information and surfacing authoritative sources [9:16]. 
Real-world agentic AI deployment outside the technology sector, illustrated through the minor league baseball example [11:08] [11:27] [11:31]. 
Notable quotes 
"We're increasingly asking AI to recommend decisions and to actually take actions. That really raises the standard for the data that's underneath it." [1:07] [1:23] 
"Context is really what transforms that information into understanding." [3:01] 
"As agents begin to recommend and execute actions, this context becomes the difference between an informed decision and just a fast, confident mistake." [3:38] 
"Access for data isn't the same as comprehension of that data. Just giving an agent access to every document in the company doesn't make it an expert any more than say giving a new employee access to every shared drive." [6:28] 
"The agent wasn't wrong, the data wasn't bad. It was reasoning very capably over information, but the information didn't reflect current reality." [11:54] 
"It's more now about having AI that can understand the business well enough to help actually drive business outcomes." [10:13] 

Planning infrastructure for the next five years right now feels a lot like packing for a vacation when the weather app says: "sunny, snowing, chance of dinosaurs." 
In this episode of The STEMINISTS Podcast, host Phoebe Goh sits down with Sarah Olibah, leader of a team of modernization solution architects at NetApp, to tackle one of the biggest challenges facing organizations today: making critical technology decisions when the only certainty is uncertainty.  
From AI initiatives that seem to appear overnight to shifting business priorities, evolving demand, and unpredictable market conditions, leaders are being asked to make infrastructure bets without the luxury of a crystal ball. Sarah explains why infrastructure planning isn't really about technology. It's about managing uncertainty and building enough flexibility to thrive when assumptions inevitably change. 
Together, Phoebe and Sarah unpack the sometimes awkward, often entertaining reality of bringing IT, finance, operations, and business leaders into the same decision-making process. Everyone wants the "right" answer, but they're often optimizing for completely different outcomes. The CIO wants agility, the CFO wants financial predictability, and the operations team wants reliability. Nobody's wrong, but nobody's speaking exactly the same language either. 
The conversation explores the growing discipline of value engineering, a framework that connects technology investments to measurable business outcomes. Sarah shares how bringing stakeholders together earlier helps organizations move beyond debates about products, pricing, and technical specifications to focus on what really matters: agility, resilience, ROI, flexibility, and long-term business value.  
In this episode, you'll discover: 
Why adaptability beats perfect forecasting every time.  
How to align technology, finance, and operations around shared business outcomes.  
The risks of designing solutions before defining priorities.  
How value engineering helps organizations navigate uncertainty with confidence. 
Why flexibility may be the ultimate competitive advantage in the AI era.  
And because no STEMINISTS episode is complete without a plot twist, Sarah closes out the conversation by sharing her latest hobby: wood whittling, including a handcrafted spoon that proves precision and patience aren't reserved for infrastructure planning alone.  
If your organization is trying to balance AI ambition with business reality, this episode offers a smart, practical, and surprisingly relatable perspective on making better decisions when the future refuses to cooperate. 
🎧 Enjoyed the conversation? Like, follow, and subscribe to The STEMINISTS Podcast. Then share this episode with your colleagues, leadership team, and fellow tech enthusiasts. After all, uncertainty may be inevitable, but navigating it is a lot easier when you're comparing notes with smart people. 
 
Learn more: 
https://www.netapp.com/data-infrastructure-insights/
 
Connect with us! 
https://www.linkedin.com/in/phoebegoh/ 
https://www.linkedin.com/in/sarah-olibah-1265b59b/ 

What happens when the technology reshaping the future is being built without enough women in the room? 
In this episode of The Steminists Podcast, host Phoebe Goh welcomes back Cecile Kellam, Senior AI Solutions Architect at NetApp and Head of AI Circles for Women in Technology (WIT), for a candid conversation about the AI participation gap and why women in tech communities may be more important than ever.  
From failing an AI-powered resume screening despite being qualified, to leading a global initiative that helps women build AI skills with confidence, Cecile brings real-world perspective to one of the biggest challenges facing the industry today. Together, Phoebe and Cecile explore how AI can amplify existing biases, why confidence often becomes a hidden barrier to participation, and what organizations can do to ensure AI is shaped by diverse voices instead of the same old perspectives.  
You'll also hear why community might be AI's secret weapon: creating safe spaces to ask the "dumb" questions, learn by doing, and influence how AI is designed, deployed, and adopted across the workplace. Because if AI is changing everything, everyone should have a seat at the table.  
And in a fun twist during "Cool and Current," Cecile shares how the latest women's health technology is finally putting women's experiences at the center of innovation.  
Because the future of AI shouldn't be trained on half the population's perspective. 
 
Key Takeaways 
AI isn't inherently fair. Without intentional inclusion, it can reinforce existing hiring and workplace biases.  
Representation matters because the people building AI directly influence the outcomes it produces.  
Confidence can be as significant a barrier as technical skills, especially when stepping into emerging fields like AI. 
Women's tech communities are evolving from networking groups into powerful engines for AI education, adoption, and influence.  
Organizations that invest in inclusive AI learning don't just do the right thing. They drive better adoption, stronger productivity, and greater business impact.  
You do not need a traditional technology background to succeed in AI. Curiosity, experimentation, and the willingness to learn can take you surprisingly far.  
The best way to close the AI participation gap might just be... using AI itself to learn faster and build confidence. 
Because the future of AI shouldn't be trained on half the population's perspective. So as always, if you enjoyed this episode, but sure to like, follow and share with your community.  
 
Learn more: https://www.netapp.com/artificial-intelligence/  
 
Learn more about WIT (Women in Technology): https://www.womenintechnology.org/  
 
Connect with us! 
https://www.linkedin.com/in/phoebegoh/ 
https://www.linkedin.com/in/%E2%98%81-cecile-kellam-%E2%98%81/  
 

While the AI world has spent the last year racing from hype cycle to reality check, this conversation with Kris Cornwall, Senior Director of Product Marketing for Enterprise Storage at NetApp, feels more timely than ever. Because it turns out that building AI is the easy part. Getting it into production, scaling it, securing it, and making it deliver actual business value? That's where things get interesting.  
AI has officially graduated from science project status. It's now expected to drive business outcomes, deliver ROI, and operate with the same resilience, security, and reliability as any mission-critical enterprise workload. But getting there is easier said than done. Kris joins Phoebe and Mekka to unpack why data remains the biggest hurdle for organizations trying to scale AI, and what it really takes to move from proof of concept to production.  
The conversation dives into one of the hottest infrastructure topics in AI today: disaggregation. Kris explains how separating performance from capacity can help organizations scale AI workloads more efficiently, maximize expensive GPU investments, and avoid the infrastructure bottlenecks that often slow innovation.  
You'll also hear how enterprise-grade storage is evolving to meet AI's growing demands, from built-in ransomware protection and cloud integration to intelligent data mobility and metadata-powered data discovery. The takeaway? AI success isn't just about choosing the right model. It's about building the right data foundation underneath it.  
Tune in for a fan-favorite conversation on enterprise AI, data infrastructure, and the unglamorous truth nobody puts in the keynote: if your data strategy isn't ready, your AI strategy probably isn't either. 
 
Learn more: 
https://www.netapp.com/artificial-intelligence/
 
Connect with us! 
https://www.linkedin.com/in/phoebegoh/ 
https://www.linkedin.com/in/kristinecornwall/ 

Everyone's obsessed with AI right now. Bigger models. Faster training. More GPUs. More data centers. More everything. But there's one small problem: none of that AI infrastructure exists without semiconductors. And the companies designing those chips are suddenly competing for the exact same resources.  
Welcome to the ultimate tech plot twist. 
In this episode of The STEMINISTS Podcast, host Phoebe Goh sits down with Diane Patton, Senior Product Manager at NetApp, to unpack how the AI boom is reshaping the silicon supply chain and putting unprecedented pressure on Electronic Design Automation (EDA), the critical technology behind modern chip design. 
As AI demand drives a global scramble for GPUs, RAM, and high-performance compute resources, chip designers are facing a new challenge. The infrastructure required to design semiconductors is becoming just as difficult to obtain as the infrastructure needed to run AI itself.  
Diane breaks down why EDA workloads are among the most demanding in enterprise IT, requiring massive compute power, low-latency storage, and the ability to process enormous volumes of data at scale. She also explains why waiting months for hardware procurement is no longer a viable strategy when innovation cycles are moving at AI speed and today's "latest and greatest" GPU can become yesterday's news before it's even installed.  
The conversation explores how semiconductor companies are turning to cloud and hybrid-cloud architectures to bypass infrastructure bottlenecks, accelerate chip development, and gain access to scalable compute resources on demand. From cloud bursting and intelligent data caching to storage tiering and AI-powered workflows, you'll hear how organizations are modernizing chip design for a world where flexibility is becoming a competitive advantage.  
Key Takeaways: 
What Electronic Design Automation (EDA) actually is 
Why AI is reshaping semiconductor infrastructure requirements 
GPU and RAM shortages facing chip design teams 
High-performance storage for EDA workloads 
Cloud vs. on-premises semiconductor design environments 
Hybrid cloud strategies for semiconductor companies 
Data management challenges in chip development 
AI-enabled opportunities for EDA workflows 
The future of semiconductor innovation and the silicon supply chain 
Whether you're a semiconductor professional, infrastructure architect, cloud strategist, or simply curious about the technology powering the AI revolution, this episode offers an insider's look at the future of chip design, EDA infrastructure, cloud computing, and the evolving silicon supply chain.
 
Learn More:
https://www.netapp.com/industries/electronic-design-automation-eda/  
https://www.netapp.com/blog/semiconductor-design-netapp-all-flash-portfolio/  
 
Connect with us!
https://www.linkedin.com/in/phoebegoh/ 
https://www.linkedin.com/in/diane-patton-7519634/