IT Infrastructure as a Conversation

Solving the $2,000 Per Hour GPU Problem With Lumen

What happens when an enterprise invests heavily in AI compute but cannot move data quickly enough to keep those expensive processors working? In this episode of Infrastructure as a Conversation, I speak with Jim Fowler, Chief Technology and Product Officer at Lumen, about why network performance is becoming one of the defining factors in enterprise AI ROI. Jim brings experience from both sides of the relationship. He previously led technology programs inside large enterprises and served on Lumen’s board before joining its management team. His description of Lumen as a logistics company for information provides a useful way to understand the network’s role. Models can process information, but the network must deliver the right data to the right location at the required speed. Our conversation begins with the $2,000 per hour GPU problem. Jim recalls speaking with a Fortune 100 CIO whose company was training its own model using data stored across multiple clouds and regions. The company had GPU availability, but moving petabytes of data to those processors had become its largest bottleneck. Expensive compute was waiting for information. We discuss why an AI workload can perform well inside a controlled development environment before slowing down when users, data sources, and cloud services become distributed. Jim says many inference workloads require response times between 10 and 20 milliseconds. Factory automation can require latency below 10 milliseconds, while natural voice conversations may need responses within five to 10 milliseconds. Poor data movement also affects employees. Jim describes spending time with data scientists who began their mornings by moving information from three global regions. The transfers took between two and two and a half hours, leaving skilled employees waiting, attending meetings, or drinking coffee before they could begin modeling. In his words, these were “idle humans” waiting on the network. The problem becomes harder as AI workloads span public clouds, private data centers, colocation facilities, and edge environments. Jim says bandwidth growth between clouds and data centers is rising far faster than traffic between premises and cloud environments. Enterprises are also managing multiple carriers, tools, and operating systems, making performance and predictability harder to control. We examine data gravity and the decision to move compute closer to data rather than transporting large data sets over long distances. Security, resilience, sovereignty, compliance, and cost must be considered according to the requirements of each workload. Jim also explains why networking is beginning to behave like a cloud service. Traditional capacity could take weeks or months to provision. A software controlled network can increase or reduce capacity as demand changes. Lumen describes this model as Cloud 2.0, with automation, APIs, and real time control replacing static capacity planning. Our discussion closes with ownership. Network, cloud, application, data, and AI teams may each measure success differently. Jim believes a business owner should oversee the complete path from data source to AI outcome because a bottleneck can appear anywhere along the way. Could the biggest constraint on your AI investment be the network connecting your data and compute? Listen to the episode and share your thoughts with me.

How EdgeBeam Wireless Uses ATSC 3.0 for One to Many Data Delivery

Could the TV towers we drive past every day provide part of the answer to rising wireless traffic and last-mile network congestion? In this episode of IT Infrastructure as a Conversation, I speak with Conrad Clemson, CEO of EdgeBeam Wireless, about using existing ATSC 3.0 broadcast infrastructure as a one-to-many wireless data delivery layer. Most modern networks are designed around individual connections. That works well when every user requests different information, but it becomes inefficient when thousands or millions of devices need the same software update, live video stream, positioning correction, public safety message, or digital signage content. Conrad explains why broadcast infrastructure offers different economics. A cellular network may deliver an identical file separately to every endpoint. EdgeBeam Wireless can use TV broadcast capacity to transmit the shared payload once across a wide area. The distinction becomes clear in connected vehicles. Conrad says some automobile manufacturers have found that cellular communication charges across the life of a lower-cost vehicle could approach the cost of the vehicle itself. Software, maps, and AI models all need regular updates, creating demand for a delivery method that does not multiply the same payload across every car. We discuss how ATSC 3.0 added data capabilities to next-generation television infrastructure. Broadcast towers already have fiber connections, transmission equipment, spectrum, and regional coverage. EdgeBeam is adding the software, control, signaling, billing, and delivery components needed to use that infrastructure for internet data. Conrad also explains why EdgeBeam is beginning with business markets such as digital signage, public safety, and positioning services, where receivers can be deployed relatively quickly. Connected vehicles, fixed wireless access, consumer video, and autonomous systems represent larger opportunities that will require broader receiver availability. Digital signage provides another useful example. A one-gigabyte content update sent individually to 1,000 screens produces one terabyte of network traffic. Broadcast distribution can send the content once, allowing screens to store a local portfolio and select the most relevant material when needed. We also consider edge computing, live sporting events, AI infrastructure, and the pressure created when identical data crosses the wireless last mile through separate connections. EdgeBeam’s approach complements cellular networks by moving suitable shared traffic elsewhere and preserving cellular capacity for interactive and personalized services. Which workloads inside your network are still being delivered through thousands of duplicate connections, and could one-to-many distribution offer a better answer? Listen to the conversation and share your thoughts with me.

Why Uptime Does Not Mean Your IT Infrastructure Is Healthy

What if the IT systems your business depends on every day are working perfectly, right up until the moment they are not? In this episode of IT Infrastructure as a Conversation, I speak with Chris Bruce, founder of Idextrus, about the hidden technical debt inside mid-sized companies, why uptime should never be confused with infrastructure health, and what IT leaders should be looking for before aging systems become an operational or security crisis. Chris has spent more than 20 years working with companies across manufacturing, wholesale, retail, consumer packaged goods, software, and e-commerce. His approach begins by understanding how technology is actually used across the business, talking with employees and examining the systems, dependencies, security controls, and processes operating behind the scenes. One of the most dangerous assumptions Chris encounters is simple: "It just works." A server may have been running for years, but if it has not been patched or properly reviewed, that does not necessarily make it stable. Companies can also have business-critical systems that nobody fully understands, infrastructure maps that no longer exist, former employees whose knowledge was never documented, shared credentials, excessive permissions, and legacy applications that everyone is afraid to touch. We discuss the different forms of technical debt that infrastructure teams need to identify, including dependency debt, credential debt, documentation debt, and years of deferred upgrades. Chris explains why documentation can become one of the biggest infrastructure risks when the only person who understands a business-critical system leaves the company. The conversation also provides a practical framework for deciding what to do with legacy technology. Should a system be modernized, integrated with newer platforms, isolated while a migration plan is developed, or finally retired? Chris explains how business value, security exposure, architecture, dependencies, and maintenance costs should influence that decision. Cloud migration is another major theme. Simply moving an existing workload from an on-premises environment into AWS, Azure, or Google Cloud does not fix the problems already inside it and can make infrastructure more expensive. Chris explains why successful modernization begins with understanding what should move, why it should move, and whether the underlying architecture needs attention first. We also examine the AI readiness gap. As businesses introduce AI agents, automation, IoT, and increasingly connected applications, weaknesses in data quality, infrastructure, security, and system architecture can become more visible. Every new integration can also create another potential attack surface. For CIOs, infrastructure leaders, IT teams, and mid-sized businesses planning cloud modernization or AI adoption, Chris offers a practical infrastructure health check. Review your external attack surface, patching processes, software lifecycle, access permissions, backups, and system dependencies. Most importantly, test whether the business can explain how its critical systems connect and what would happen if one of them failed. This conversation is a reminder that infrastructure resilience is not measured by how long a system has managed to stay online. The better question is whether you understand what is running, who has access to it, what depends on it, how quickly you could recover it, and whether the infrastructure you have today is ready for what the business wants to do tomorrow.

How Gi21 Capital Sees The Next Wave Of AI Infrastructure Growth

What does it really take to build the infrastructure powering the AI economy? While much of the conversation around artificial intelligence focuses on models, applications, and breakthroughs, far less attention is given to the physical foundations making it all possible. Behind every AI workload sits an enormous amount of infrastructure, from power generation and cooling systems to land acquisition, grid connectivity, and data center design. In this episode of IT Infrastructure as a Conversation, I speak with Damir Špoljarič, a technology entrepreneur, investor, and infrastructure specialist whose journey began at just 17 years old when he founded VSHosting. Over the following two decades, he helped grow the company into one of Central Europe's leading cloud providers while building a data center that achieved something few facilities can claim, verified 100% uptime for more than a decade. Damir shares the lessons learned from designing for resilience at a level where failure simply isn't an option. He explains why many operators underestimate the importance of redundancy, how early decisions around infrastructure design can have consequences years later, and why reliability often comes down to planning for scenarios that may never happen. We also discuss how AI is changing the economics and engineering of modern data centers. As compute density continues to rise, traditional approaches are being pushed to their limits. Damir explains why liquid cooling is becoming increasingly important, how power requirements have changed dramatically, and what operators must consider when designing facilities capable of supporting next-generation AI workloads. The conversation also turns to Europe's growing demand for AI compute capacity and the challenges involved in bringing new facilities online. From securing grid connections and navigating lengthy permitting processes to finding suitable locations with access to affordable energy, Damir offers a behind-the-scenes look at obstacles that rarely make the headlines but shape the future of digital infrastructure. We also explore digital sovereignty, sustainability, renewable energy, and why waste heat from data centers may become an overlooked opportunity for local communities. Along the way, Damir shares his thoughts on robotics, long-term infrastructure investment, and why he believes demand for AI resources is still in its earliest stages. If you've ever wondered what sits beneath the AI services we use every day, this conversation offers a fascinating look at the engineering, investment, and strategic planning required to build the infrastructure supporting the next generation of technology. What role do you think Europe should play in building the infrastructure needed for the AI era, and are we moving quickly enough to meet future demand?

From SD-WAN to AI Traffic: How Enterprise Networks Are Evolving

What happens when AI workloads begin to overwhelm the network infrastructure originally designed for human browsing and SaaS consumption? In this episode of IT Infrastructure as a Conversation, I’m joined by Jamie Pugh from Globalgig to discuss why enterprise connectivity is rapidly becoming one of the biggest blind spots in the AI era. While much of the industry conversation focuses on GPUs, models, and data centers, Jamie explains why the network itself is now under growing pressure from entirely new traffic patterns driven by AI systems communicating with other AI systems. We explore how enterprise infrastructure was largely built around human behavior, employees accessing applications, downloading files, and consuming cloud services. AI changes that model completely. Today, agents are constantly interacting with tools, inference engines are querying massive data stores, and cloud environments are exchanging huge volumes of east-west traffic across regions in real time. Jamie explains why many SD-WAN architectures and broadband-heavy deployments were never designed for these sustained, burst-heavy workloads. The conversation also examines the growing importance of cloud on-ramps and why many organizations discover bottlenecks only after deploying AI-enabled services into production. Jamie shares how asymmetric broadband connections, fragmented carrier relationships, and static connectivity models can quietly introduce latency, resilience, and observability problems that directly impact AI performance and user experience. One of the most interesting parts of the discussion centers on how dependent modern workflows are becoming on AI tools. Jamie talks candidly about using platforms like Claude, Perplexity, and ChatGPT throughout his working day and why losing connectivity now feels less like a temporary inconvenience and more like losing access to an essential member of the team. That shift in expectation is forcing infrastructure leaders to rethink resilience, automation, and real-time observability across hybrid and multi-cloud environments. We also discuss programmable networks, predictive routing, network-as-a-service fabrics, and the growing move toward centralized control planes that can dynamically adapt to changing AI traffic patterns. Jamie explains why enterprises need to stop thinking purely about north-south traffic and start preparing for a future dominated by east-west communication between clouds, data centers, agents, and inference platforms. There is also a valuable conversation around governance, security, and data sovereignty as organizations increasingly bring AI inference closer to private infrastructure rather than relying entirely on public models. Jamie argues that networking, security, and AI strategy teams can no longer operate in silos if businesses want to scale AI safely and effectively. If your organization is building toward an AI-first future, this conversation offers a timely look at the infrastructure challenges many enterprises are only beginning to recognize.

Syndigo on Why AI Commerce Is Failing Without Better Product Data

What if the biggest problem in AI-powered commerce isn’t the AI at all, but the data feeding it? In this episode of IT Infrastructure as a Conversation, I spoke with Tarun Chandrasekhar, Chief Product Officer at Syndigo, about the hidden infrastructure powering modern commerce and why product data has suddenly become one of the most strategic assets inside every retail and consumer brand. As AI shopping assistants, conversational commerce, and agentic retail experiences rapidly move into the mainstream, many companies are discovering a hard truth. Their product information systems were never built for an AI-first world. Tarun explained why decades of fragmented product records, disconnected systems, inconsistent metadata, and siloed workflows are now becoming major blockers to reliable AI-driven discovery and personalization. This episode offers a fascinating look at why “single source of truth” projects continue to fail across enterprises decades after organizations first started chasing them. Tarun argued that this is less a technology problem and more a people-and-process problem, where organizational handoffs and disconnected ownership models continue to create friction across data pipelines. We also explored the rise of agentic commerce, AI readiness scoring for enterprise data, and how companies are now being forced to treat product data as infrastructure rather than simply marketing content. Tarun shared how smaller brands sometimes leapfrog larger enterprises by moving faster, adopting AI-native workflows more easily, and avoiding decades of technical debt. We also discussed Syndigo’s acquisition of OneWorldSync and how ratings, reviews, and product syndication data are increasingly interconnected within AI-powered commerce ecosystems. The long-term vision is a world where product data continuously improves through feedback loops between customers, retailers, AI systems, and manufacturers. If you work in retail technology, AI infrastructure, enterprise data management, supply chain systems, or digital commerce, this episode offers a valuable behind-the-scenes look at the systems quietly shaping the future of how products are discovered, trusted, and purchased online. Useful Links Syndigo Acquires 1WorldSync to Lead AI-First PXM Syndigo Connect with Syndigo’s Chief Product Officer, Tarun Chandrasekhar

• Transcript

Why Infrastructure Needs A Survivability Layer: HyperBUNKER And The Shift To True Offline Recovery

In this episode, I’m joined by Imran Nino Eškić and Boštjan Kirm from HyperBUNKER, two leaders whose perspective has been shaped by more than 50,000 real-world data loss and ransomware cases. This is not a conversation about theoretical security models or another incremental backup feature. It’s a discussion about what actually survives when production systems, identity layers, and cloud replicas have all been compromised. For years, infrastructure has been designed around availability, scale, and performance. Recovery was treated as a process that would work when needed. But as attackers have grown more patient and methodical, they now target recovery paths first, quietly mapping environments and neutralising backup systems long before an incident becomes visible to the business. That shift forces a new architectural question for infrastructure leaders. Where is the layer that remains reachable when everything connected has been taken down? We explore why so many environments that claim to be air-gapped or immutable still rely on credentials, control planes, and automation, and how those dependencies create hidden single points of failure. Imran and Boštjan explain how HyperBUNKER introduces a physically isolated survivability layer into modern infrastructure, using a hardware-enforced, one-way ingestion process and a double air-gap design that removes the network from the vault entirely. No IP address, no inbound ports, and no authentication surface to attack. This leads to a wider conversation about infrastructure governance, cyber insurance, and regulatory pressure. Insurers are increasingly focused on whether a final, untouchable copy of critical data exists, because the largest financial losses now come from failed recovery rather than the initial breach. That reality is pushing offline recovery out of the basement and into board-level architecture discussions. We also tackle the practical challenge every organisation faces. If only a small percentage of data can be placed in a fully isolated vault, how do you decide what keeps the business alive? That decision, as we discuss, cannot sit with IT alone. It requires operational and executive alignment around what the company must have to restart after a catastrophic event. This episode reframes resilience as an infrastructure design principle rather than a security feature. It asks where a survivability layer should sit alongside cloud platforms, backup software, and existing controls, and why the future of Infrastructure as a Service may depend as much on guaranteed recovery as it does on uptime. If your architecture assumes that recovery will always be there when you need it, this conversation may change how you think about your entire stack.

How To Simplify Storage, Virtualization, And Recovery At Scale

What happens when the infrastructure your business depends on becomes too complex, too expensive, or too slow to recover when something goes wrong? In this episode of Infrastructure as a Service, I sit down with Tvrtko Fritz to talk about how EuroNAS has evolved from “NAS for the masses” into a platform helping organizations simplify storage, virtualization, backup, and scale-out infrastructure. What stood out to me in this conversation was how much of that journey was shaped by real customer pain, from complex deployments to the growing pressure of managing modern environments without a team of specialists. We also get into one of the biggest infrastructure talking points right now, the VMware shift. Tvrtko shares why many customers are not moving because they want to, but because they feel they have to, and how EuroNAS is helping reduce that friction with migration support, VM import tools, and a more predictable licensing model. It is a practical look at what organizations are really facing when they need a plan B but cannot afford disruption. Another part of the conversation I found especially interesting was how recovery speed is becoming just as important as backup itself. Tvrtko explains how instant backup and recovery can change the experience from hours of downtime to seconds, whether that means restoring a full virtual machine or pulling back a single file. We also talk about simplifying Ceph deployments, reducing setup times from days to minutes, and why infrastructure teams increasingly need solutions that let them focus on applications and outcomes rather than wrestling with storage architecture. If you are rethinking your virtualization strategy, looking for more predictable infrastructure costs, or trying to understand how to make enterprise storage less painful to manage, this episode is packed with useful insights. After listening, do you think the future of infrastructure belongs to platforms that hide complexity rather than expose it, and what would make you confident enough to make a switch? Share your thoughts.

PuppyGraph at IT Press Tour: Zero-ETL Graph Analytics on Your Existing Data

What does “infrastructure” mean when your data stays exactly where it is, yet suddenly behaves like a graph? I met Weimo Liu, CEO and co-founder of PuppyGraph, during an IT Press Tour presentation, and I wanted to bring his story to Infrastructure As A Conversation because this is a data infrastructure conversation at its core. Weimo’s pitch is simple to say and harder to pull off: keep a single copy of data in your lake or warehouse, skip the ETL pipelines, and still run graph queries with subsecond performance. Weimo’s background explains why this is more than a clever demo. He worked at TigerGraph, then on Google’s F1 team, and PuppyGraph sits right between those worlds. In our conversation, he walks me through how they treat graph queries as a set of node and edge operations that can be optimized, parallelized, and evaluated in a vectorized way, which is how they keep performance predictable when workloads get real. We also get into the practical details infrastructure teams care about. PuppyGraph is a read-only engine, which changes the trade-offs around concurrency, governance, and operational risk. Instead of copying data into a separate graph store and building a second set of controls, you can query relationships where the data already lives, then write results back into the lake for other engines to consume. The upside is simpler architecture and less duplication. The compromise is that you are not getting transactional graph updates, and Weimo is clear about why that is acceptable for the OLAP-style workloads his customers run. From there, the use cases start to make sense fast. Cybersecurity teams with logs sitting in object storage, fraud detection scenarios where latency matters, and internal AI chatbots that struggle with too many tables and brittle SQL generation. Weimo has a sharp analogy for that last part, text-to-graph queries behave more like a train on rails, which can help AI stay inside defined relationships and reduce messy answers. If you are building modern data platforms and you are tired of pipelines multiplying, this episode is a thought-provoking look at what happens when graph analytics becomes a query layer rather than a destination system. And it all started with a dog-themed name and a surprisingly cheap domain.

Why Infrastructure Complexity Is Still Being Underestimated

When does automating tasks stop being enough, and when does infrastructure itself need to become a shared conversation across teams? In this episode of IT Infrastructure as a Conversation, I’m joined by Peter Sprygada, Chief Architect at Itential, for a deep and refreshingly pragmatic discussion about how infrastructure operations have evolved from scripts and isolated automation into something far more complex and interconnected. Peter has spent more than a decade working across enterprise networks, cloud platforms, and automation tooling, and that experience shows in how he talks about what has genuinely changed, and what has not. We trace the journey from early network automation, born out of engineers trying to escape repetitive CLI work, to the point where automation alone starts to break down. Peter explains why automation excels in domains but struggles across end-to-end systems, and why orchestration becomes essential once infrastructure has to align with real business intent. Instead of teams pointing fingers when something fails, orchestration creates a common language that allows network, cloud, application, and platform teams to work toward the same outcome, even when they use different tools and terminology. We also tackle AI head-on, separating operational reality from conference stage promises. Peter shares his own initial skepticism, why treating AI as a tool rather than a silver bullet matters, and how the same lessons learned from automation apply again today. We talk about governance, guardrails, and what Peter calls the boring stuff, the logging, security, and controls that actually make innovation sustainable at scale. As infrastructure complexity continues to rise, he argues that many leaders still underestimate just how much engineering effort is required to keep modern platforms reliable. Looking ahead, Peter outlines how the traditional boundaries between orchestration, automation, and observability are already starting to blur, and why investing in platforms that can evolve matters more than chasing the latest shiny technology. This conversation stays grounded in real operational challenges, real tradeoffs, and real lessons learned in the field, not abstract futures. As infrastructure stacks continue to grow more complex and AI becomes part of daily operations, are IT leaders ready to treat infrastructure less like a collection of tools and more like an ongoing conversation that never really stops? Useful Links Connect with Peter Sprygada Learn more about Itential, Thanks to our sponsors, Alcor, for supporting the show.

Why the Endpoint Is Becoming the New Control Point for IT Leaders

What happens when someone who helped shape the early days of virtualization returns to the field to lead a new era of secure, flexible computing? At IGEL’s Now & Next event in Frankfurt, I sat down with Peter Goldbrunner, Regional Vice President for Central Europe, to explore how his decades of experience at Citrix and Nutanix are helping drive IGEL’s transformation into a software-focused security leader. Peter’s background provides him with a unique vantage point on the evolution of enterprise technology. From server-based computing to cloud and now prevention-first endpoint security, he has witnessed every wave of IT transformation firsthand. In this conversation, he explained how IGEL’s evolution is meeting the moment for customers facing economic pressure, complexity, and the demands of hybrid work. He also described why the endpoint is emerging as the new control point for business continuity and security, rather than the weakest link. We also discussed what it means to lead transformation within a company that has already reinvented itself. Peter shared how the German-speaking markets are balancing cost control with innovation, why local trust matters, and how IGEL’s role-based approach to security helps organizations prepare for what comes next. His perspective on collaboration, learning, and execution offered a clear reminder that transformation is not just about technology; it is about people, priorities, and timing. As the Now & Next tour continues, Peter’s insights demonstrate why resilience and prevention are now inseparable for any organization that wants to thrive in a world where security and productivity are inextricably linked. How can companies ensure their IT strategies are ready for what comes next? Useful Links Connect with Peter Goldbrunner, on LinkedIn Learn more about IGEL Follow on LinkedIn, Twitter and YouTube Tech Talks Daily is Sponsored by NordLayer: Get the exclusive Black Friday offer: 28% off NordLayer yearly plans with the coupon code: techdaily-28. Valid until December 10th, 2025. Try it risk-free with a 14-day money-back guarantee.

Observability without Overload: Lessons from Chronosphere’s Field CTO

In this episode of IT Infrastructure as a Conversation, I sit down with Bill Hineline, Field CTO at Chronosphere, to talk about something every IT leader has wrestled with at some point, observability. Bill has spent more than 25 years in the industry, including leading observability at United Airlines, so he knows firsthand what happens when data collection turns from helpful to overwhelming. We talked about why the old collect-everything mindset has created what Bill calls data landfills, how organizations can shift toward collecting data with intent, and why so many teams still struggle with alert fatigue and burnout. Bill also shared his take on how to build observability around business outcomes rather than dashboards, and why connecting metrics to real customer value is the only way to truly measure ROI. What really struck me was Bill’s perspective on culture. He believes that observability should be treated as a first-class citizen, something that empowers teams rather than burdens them. Moving from reactive firefighting to proactive, data-driven decision-making is not easy, but as Bill explains, it starts with asking one simple question: why are we collecting this data in the first place? If you have ever been woken up at 3 a.m. by an alert that turned out to be nothing, this conversation will feel painfully familiar. But it is also full of practical insights on how to change that story for good. Sponsored by NordLayer: Get the exclusive Black Friday offer: 28% off NordLayer yearly plans with the coupon code: techdaily-28. Valid until December 10th, 2025. Try it risk-free with a 14-day money-back guarantee.

• Transcript

Phison Electronics and the Future of AI-Native Infrastructure

In this episode, I’m joined by Michael Wu of Phison Electronics, recorded shortly after our meeting on the IT Press Tour in Silicon Valley. Michael takes us inside Phison’s latest breakthrough: the aiDAPTIV+ platform and its integration with StorONE’s ONEai solution. Together, they’re reshaping how enterprises think about AI training, inference, and data sovereignty. Michael explains how aiDAPTIV+ acts as expansion memory for GPUs, reducing power consumption and cutting hardware costs by up to 10x. We also dig into the partnership with StorONE, which has produced a plug-and-play, storage-based AI solution that makes large language model training accessible to organizations of all sizes—including smaller businesses and universities that traditionally struggle with GPU access. From the launch of the E28 Gen5 AI-enabled SSD controller to the endurance-driven Pascari X200Z SSDs, Michael shares the technical innovations under the hood and what they mean for performance and reliability. He also looks ahead to future workloads, where models with trillions of parameters will demand smarter, more scalable storage architectures. If you’re an IT leader weighing the trade-offs between cloud-based AI and secure, on-premises solutions, or you’re simply curious about how storage is becoming central to AI acceleration, this conversation will give you a fresh perspective. ********* Visit the Sponsor of Tech Talks Network: Land your first job in tech in 6 months as a Software QA Engineering Bootcamp with Careerist https://crst.co/OGCLA

Cohesity’s Infrastructure Playbook for Cyber Recovery and Insights

How can organizations protect their most valuable asset, data, while unlocking its full potential through AI-driven insights? That is the question I explored on the IT Press Tour in Silicon Valley during a face-to-face conversation with Sanjay Poonen, President and CEO of Cohesity. In this episode, Sanjay shares how Cohesity evolved from reinventing backup and recovery to leading the market in data security and cyber resilience. We discuss the game-changing acquisition of Veritas’s NetBackup business, the company’s growing footprint across healthcare, finance, government, and retail, and why uniting cultures and customers is key to their next chapter. We dive into how Cohesity is using AI to transform backup data into a source of real-time intelligence, including its patented Retrieval Augmented Generation (RAG) capabilities developed in partnership with NVIDIA. Sanjay also explains how this innovation is enabling businesses to meet strict data sovereignty requirements while benefiting from cloud agility. From cyberattack recovery and on-premises innovation to working with some of the biggest names in AI and cloud, Sanjay offers a candid look at leadership in a time of rapid growth. If you want to understand how AI and cybersecurity are converging to shape the future of enterprise data, this conversation delivers practical insight, strategic thinking, and a clear vision for what is ahead.

Can Estonia Compete with AWS? Storadera’s Tommi Kannisto Thinks So

In this episode of Infrastructure as a Conversation, I’m joined by Tommi Kannisto, founder of Storadera, a cloud storage company based in Estonia that’s quietly building a smarter, simpler alternative to hyperscalers like AWS. We dive into how Storadera has engineered its own storage software from scratch to deliver secure, S3-compatible cloud storage with a unique hyper-converged architecture. It’s all about cutting unnecessary hardware, avoiding bottlenecks, and delivering transparent pricing that makes sense to growing businesses. Tommi explains: How Storadera’s hyper-converged design replaces gateways and load balancers with lean software Why performance with small files became a key differentiator What makes their multi-tenant system attractive to retail partners The impact of data sovereignty concerns on customer growth across Europe Why Canada is now looking eastward, not southward, for storage partners The Estonian tech culture that helped birth 10 unicorns from a country of just 1.3 million We also talk about the cultural mindset that powers Estonia’s startup scene, from engineers cold-messaging CEOs for advice to a national infrastructure designed for digital innovation. Tommi shares Storadera’s future roadmap, including plans to use AI to optimize disk read and delete operations without raising prices. If you’re curious about what comes after hyperscale, why storage software still matters, or what makes Estonia such a hotbed for digital infrastructure innovation, this is an episode you’ll want to hear. 🔗 Learn more at storadera.com 📍 Want Storadera in your country? Join their regional waitlist on the website.

Simplifying Stateful Workloads: The Rise of Kubernetes-Native Data Platforms

What if running databases in Kubernetes could be as simple as spinning up a container—without cloud lock-in or complexity? In this episode of IT Infrastructure as a Conversation, I’m joined by Tamal Saha, founder of AppsCode, a company rethinking how we manage data on Kubernetes. We met during the IT Press Tour in London, and this conversation dives deep into how AppsCode is tackling one of the most stubborn challenges in enterprise IT: stateful workloads. From CubeDB to Stash, Voyager, and KubeVault, Tamal walks us through the full story—from his early days at Google and the emergence of Kubernetes to bootstrapping a company through open source tools and evolving it into a full-fledged enterprise platform. We explore: The challenges of running databases in Kubernetes and why traditional VM-based infrastructure falls short Why database provisioning, backups, secret management, and ingress need Kubernetes-native solutions How AppsCode pivoted from open source tools to a sustainable business model Real-world enterprise use cases—including a major European telco’s cloud-native transformation The road ahead: vector databases, open telemetry, and AI-driven automation If you're a platform engineer, DevOps leader, or just curious about where Kubernetes is headed next, this conversation offers rare insights into building data platforms from the ground up, with a practical, product-led mindset. So tune in to hear how one founder turned a container-native vision into a global business that’s helping companies modernize data operations without losing control.

On-Prem, In Control: The Infrastructure Shift Toward Data Sovereignty

In this episode of IT Infrastructure as a Conversation, I sit down with Thomas Bak, CEO of Auwau, fresh from our meeting at the IT Press Tour. We explore how his journey from running a managed service provider to founding Auwau led to the creation of CloudTility — a self-service, multi-tenant portal built for modern data protection and storage needs. Thomas unpacks how CloudTility helps MSPs and enterprises unify control across vendors like IBM, Rubrik, Cohesity, and more. He explains why customers are leaning toward on-premise deployment to maintain control and meet compliance demands, and how the platform supports complex organizational structures with flexible role-based hierarchies and automated billing. We also cover how CloudTility empowers IT teams to act more like internal MSPs, reduces reliance on spreadsheets, and enables partners to white-label services with minimal overhead. Whether you’re managing backups across regions or looking to scale your services, Thomas shares practical insights into solving real infrastructure challenges with clarity and control.

Fixing Legacy Data: How Quesma Reinvents Database Migrations and Insights

On this episode of IT Infrastructure as a Conversation, I explore a fresh approach to one of the oldest headaches in enterprise IT: migrating legacy databases without breaking everything. My guest is Jacek Migdał, co-founder and CEO of Quesma, a startup tackling the messy reality of old data stacks, rigid licensing, and costly, high-risk migration projects. Jacek shares how Quesma’s database gateway acts as a smart proxy, allowing companies to switch data stacks gradually, test changes safely, and avoid the dreaded “big bang” migration that so often fails. We unpack how Quesma blends pragmatic engineering with AI-driven automation, from SQL extensions that enrich data inside the database to “smart charts” that generate meaningful visualizations without complex BI tools. Jacek also explains why even modern industries like telecom and travel still wrestle with legacy systems and how a flexible, proxy-based approach keeps critical operations online while modernising behind the scenes. If your team is wrestling with outdated data infrastructure but cannot afford downtime, you will want to hear how Quesma turns risky transitions into manageable, incremental improvements. This is a candid look at the reality behind today’s data stack promises and a reminder that when it comes to enterprise infrastructure, practical steps often beat grand plans.

Fabrix.ai and the Future of Agentic AI for Enterprise IT

In this episode of IT Infrastructure as a Conversation, recorded live at the IT Press Tour, I caught up with Raju Datla, CEO of Fabrix.ai, to talk about a shift that could redefine how IT operations are managed. Formerly known as CloudFabrix, the company has evolved with a sharper focus on what it calls agentic AI: technology that works alongside humans to make smart, controlled decisions at scale. Raju’s story begins at the Indian Institute of Technology and winds through Silicon Valley, where he has founded several ventures grounded in solving real-world tech problems. Reducing Noise, Increasing Value One of the standout achievements we discussed is Fabrix.ai’s ability to reduce alert noise by up to 95 percent. In large environments with millions of daily notifications, that kind of reduction changes how teams work. Instead of chasing false alarms, IT professionals can focus on what matters: stability, uptime, and real outcomes for the business. The platform does this through a layered architecture Raju describes as the three fabrics: data, AI, and automation. Each plays a role in bringing clarity and action to complex infrastructure environments. Data is unified from dozens of sources. AI makes decisions based on context. Automation executes those decisions while keeping humans involved in key steps. Strategic Moves and Trusted Partners Fabrix.ai has not gone it alone. Through close relationships with Cisco, IBM, and Splunk, the company has stayed connected to both market demand and enterprise pain points. These partnerships are not just logos on a slide. They are part of how the platform has been built to handle real-world complexity. And the results are tangible. Whether it is automating resolution, tracking full alert lifecycles, or offering visual storyboards for better decision-making, Fabrix.ai is helping enterprise teams keep up with a pace of change that is not slowing down. Agentic AI in Practice The concept of agentic AI comes up often in this conversation, and for good reason. Unlike systems that simply follow rules or surface alerts, this approach blends autonomy with awareness. It does not just generate insights; it acts on them. And it does so in ways that respect the role of human judgment. Raju explains that this is not about removing people from the loop. It is about giving them systems that can scale, adapt, and support smart decisions. In that sense, Fabrix.ai is not replacing IT teams. It is extending what they can do. For leaders wrestling with fragmented tools, alert fatigue, and growing complexity, this episode offers a fresh perspective and a reminder that practical, scalable AI is already here. Raju’s parting advice to entrepreneurs and IT leaders alike? Solve the problems you care about. Passion always carries more weight than a quick exit plan. Listen in to learn how Fabrix.ai is helping enterprises bring order to operational chaos, one intelligent decision at a time.

Is AI Infrastructure Broken? A Candid Conversation with Volumez

Is AI Infrastructure Broken? A Candid Conversation with Volumez AI adoption is accelerating, but most enterprises are still stuck in the pilot phase. Cloud costs keep climbing, GPUs go underutilized, and data pipelines struggle to keep pace. If AI is the future, why is the infrastructure built to support it so often stuck in the past? In this episode, recorded live in Silicon Valley during the IT Press Tour, I sit down with John Blumenthal, Chief Product Officer at Volumez, and Diane Gonzalez, Senior Director of Business Development and Product. Together, we unpack what is really holding AI back and explore how Data Infrastructure as a Service (DIaaS) could change the equation. We explore: Why traditional AI infrastructure models are inefficient and unsustainable How DIaaS enables just-in-time, automated infrastructure tuned to each workload The role of GPU and data scientist efficiency in determining AI ROI How Volumez achieved industry-leading results in the MLCommons benchmark Why hybrid and multicloud strategies demand a fundamentally different infrastructure approach John and Diane share firsthand insights from working at the intersection of data, cloud, and AI infrastructure. They argue that achieving meaningful return on AI investment requires more than hardware upgrades or clever provisioning. It means embracing automation, profiling cloud capabilities in real time, and architecting pipelines that adapt to the specific demands of each phase in AI and ML workflows. Whether you're building AI platforms, running data science teams, or managing cloud infrastructure, this conversation offers a grounded look at how to make AI actually scalable. Are you wasting your most valuable resources or are you ready to run AI workloads at full speed with none of the bloat?