Enterprise AI Infrastructure
AI Infrastructure Keynote Speaker
Mark Lynd is the AI infrastructure keynote speaker for executives who have to fund, build, and operate AI at enterprise scale, and who are tired of stage talks that stop at the GPU and never reach the power, water, network, and operating-cost decisions that actually determine whether the program survives the next budget cycle. Mark is a 5x CIO/CISO and one of a small number of practitioners ranked Top 10 globally in 5 Thinkers360 disciplines simultaneously: #3 Data Center, #4 Cloud, #4 Security, #5 Cybersecurity, and #7 Artificial Intelligence (and #1 globally in Cybersecurity in 2023). The Top 10 spans in Data Center and Cloud are the operating signal that matters for AI infrastructure audiences. Mark is currently Head of Executive Advisory & Strategy at Netsync, and the AI infrastructure conversations he is in this quarter are not theoretical. They are CEOs, CFOs, CIOs, and CTOs sitting in front of a multi-year AI infrastructure plan and trying to make it real. He has spoken on AI infrastructure at industry main stages, corporate offsites, and government forums, and he is one of a small number of public voices Thinkers360 has named to its Top 10 Global Thought Leaders on AI Infrastructure list for 2026. AI infrastructure audiences leave a Mark Lynd keynote with the full operating picture: the GPU and accelerator choice, the data-center power and cooling envelope, the water reality, the network fabric for AI training and inference traffic, the storage architecture for training data and model weights, the operating cost that hits six months after the model goes live, the security perimeter around the AI compute estate, and the people and partner decisions that decide whether the program scales or stalls. The frameworks audiences take back include The AI Adoption Tipping Point Model, which gives infrastructure leaders a four-stage view of where the enterprise actually is on AI (Experiment, Pilot, Embedded, Load Bearing), and The Enterprise AI Trust Score, which puts a board-ready frame on how AI infrastructure choices feed AI governance outcomes. Industries Mark currently advises and speaks to on AI infrastructure include financial services and banking, healthcare and life sciences, manufacturing and industrials, energy and utilities, technology and software, public sector and government, higher education, and the hyperscaler partner ecosystem. The questions are the same in every room: how do we size the AI infrastructure footprint without overbuilding, where does the power actually come from, where does the water actually go, what does the network for AI traffic look like, what is the real operating cost when the inference bill stabilizes, where do we host (own data center, colo, hyperscaler, neo-cloud, regional sovereign cloud), how do we govern shadow AI compute purchased outside the formal stack, and how do we build a multi-year AI infrastructure plan the board will fund. Mark gives audiences the operating model, the named frameworks, and the practitioner examples from this quarter, so they leave with the picture they came for and the slide deck they can use on Monday.
Live on stage · International keynote
Quick Answer
Mark Lynd is an AI infrastructure keynote speaker, 5x CIO/CISO, and Thinkers360 Top 10 globally in 5 disciplines (Cybersecurity, AI, Cloud, Security, Data Center). He is one of a very small group of speakers who can address AI compute, AI data centers, AI networking, AI power and water, and the operating cost of AI from inside the practitioner's chair, with named frameworks audiences take back to their teams.
Last updated: April 2026 · Verified by Mark Lynd, 5x CIO/CISO with 25+ years of experience
Keynote Topics
The AI Infrastructure Operating Picture: Compute, Power, Water, Network, Cost
Most AI infrastructure conversations get stuck on GPUs. Mark walks the audience through the full operating picture, AI compute and accelerator choice, data-center power and cooling envelope, water budget, training and inference network fabric, storage for training data and model weights, and the operating bill that lands six months after the model goes into production. The practitioner view executives need to fund the right plan.
Best for: CEO, CFO, CIO, CTO audiences, AI infrastructure summits, hyperscaler partner events
Duration: 45-75 minutes
Sizing AI Infrastructure Without Overbuilding
Half of the enterprise AI infrastructure plans Mark sees are over-engineered for the workload the business will actually run. The other half are under-engineered for what comes in year two. Mark covers the sizing model he uses in advisory work, the unit economics, the workload-mix assumptions, and the elasticity calls that decide whether the plan still holds at the end of the budget cycle.
Best for: Enterprise architects, AI platform leaders, capacity planners, CFO partner events
Duration: 45-60 minutes
Power, Cooling, and Water: The Real AI Infrastructure Constraints
Power availability now decides where AI infrastructure can be built before any technology choice is made. Water consumption is the new community and ESG conversation. Mark covers the power-and-cooling reality (liquid cooling, immersion, direct-to-chip), the water disclosure conversation regulated industries are walking into, and what the leading operators are doing to keep the AI program on the schedule the board approved.
Best for: Utility forums, data-center industry events, sustainability conferences, regulated industry C-Suite
Duration: 45-60 minutes
Where to Host AI: Own, Colo, Hyperscaler, Neo-Cloud, Sovereign
The hosting decision for AI workloads is not the same hosting decision the enterprise made for the rest of the application portfolio. Mark walks audiences through the five hosting modes that matter now (own data center, colocation, hyperscaler, neo-cloud, regional sovereign cloud), what each is actually good at, what each costs once the inference traffic stabilizes, and how leading enterprises are blending them.
Best for: CIO summits, cloud strategy forums, hyperscaler partner events, sovereign-cloud audiences
Duration: 45-60 minutes
The Network Fabric for AI: Training Traffic, Inference Traffic, East-West Reality
AI workloads break network designs that were sized for a north-south web traffic world. Mark covers the AI network fabric reality, training east-west traffic patterns, inference traffic at the edge, the move to Ethernet for AI versus InfiniBand, and the unsexy work of redesigning the data-center network so the AI program does not get bottlenecked by the spine.
Best for: Network operations leadership, AI platform engineering forums, data-center events
Duration: 45-60 minutes
Securing the AI Compute Estate: Identity, Perimeter, Supply Chain
AI infrastructure is a security perimeter most organizations have not yet treated as one. Mark covers the security work the practitioner sees today, identity-for-AI workloads, model and weight protection, the supply chain for training data and pretrained models, and the security perimeter around the GPUs themselves. Built on Mark's CISO operating history and his work on the AI plus cybersecurity intersection.
Best for: CISO summits, AI security forums, government cybersecurity events
Duration: 45-60 minutes
Shadow AI Compute: The Quiet Capacity Problem
Shadow AI is usually framed as a data leakage story. The infrastructure side of it is a quiet capacity problem, lines of business spinning up paid AI usage outside the formal stack, with no view of cost, identity, security, or vendor concentration. Mark covers what shadow AI compute actually looks like in production environments today and the executive moves that bring it back into the plan.
Best for: CIO summits, CFO forums, AI governance events, IT operations leadership
Duration: 45-60 minutes
The Multi-Year AI Infrastructure Plan the Board Will Fund
Every enterprise above a certain size is now building a multi-year AI infrastructure plan and presenting it to a board. Most are not landing. Mark walks audiences through the structure that does land, the unit economics the board is going to ask about, the risk surface the audit committee is going to read, and the adoption velocity the strategy committee wants to track. Anchored on The AI Board Briefing Triangle.
Best for: Board retreats, audit-committee briefings, CEO-CFO offsites, multi-year-plan forums
Duration: 30-60 minutes
Most AI infrastructure talks stop at the GPU. The decisions that actually matter happen at the power contract, the water budget, the network fabric, and the operating bill that lands six months after the model goes live. That is the conversation I bring to the stage.
, Mark Lynd, 5x CIO/CISO, Head of Executive Advisory & Strategy at Netsync
Why Organizations Choose Mark as Their AI Infrastructure Keynote Speaker
Top 10 globally in 5 Thinkers360 disciplines simultaneously, including Data Center (#3), Cloud (#4), and AI (#7). The single most direct signal that Mark can hold the full AI infrastructure conversation, compute, hosting, network, and the AI strategy that sits on top of it.
Named to Thinkers360 Top 10 Global Thought Leaders on AI Infrastructure for 2026. A current, public recognition specifically in this category.
5x CIO/CISO with operating responsibility at $72B financial-services scale. Real budget, real outages, real audit committee.
Head of Executive Advisory & Strategy at Netsync. Currently sitting with CEOs, CFOs, CIOs, and CTOs on multi-year AI infrastructure plans every week.
Practitioner frameworks audiences take home. The AI Adoption Tipping Point Model, The Enterprise AI Trust Score, The AI Board Briefing Triangle.
Speaks the full stack. Compute, hosting, network, power, water, security perimeter, and the operating cost that hits after launch. Not a one-layer talk.
Carrier-friendly and industry-respectful. No vendor pitch from the stage. No fear pitch. No hype on AI capability the audience has to discount the next morning.
Author of three published books on cybersecurity, cyber insurance, and a 72-hour critical-infrastructure scenario. Citable, durable points of view.
Cybervizer newsletter, weekly writing on the AI and cybersecurity intersection that AI infrastructure audiences increasingly read for the governance side of the picture.
100+ keynotes delivered, with main-stage experience at RSA, Oracle CloudWorld, Dell Technologies World, IBM Think, Cisco Partner Summit, and the Technology Ball.
Speaker Reel
Watch Mark on Stage
A sampling of keynotes, panels, and live broadcast appearances , RSA, Oracle CloudWorld, Dell Technologies World, ESPN College Football Awards, and the Technology Ball.
What Audiences Say
Feedback From Event Hosts and C-Suites
Mark stands apart. His credibility isn’t rooted in a title from years ago , it’s built through the work he’s doing every day in the field. When he speaks about our technology, enterprise buyers pay attention because they know his perspective is grounded in real-world experience.
Shira Rubinoff
CEO, The Cybersphere Group
Mark delivers more than a presentation , he delivers operational insight from the front lines. Instead of theory, he shares what is actually working in real environments. Our audience of CISOs and security leaders left with practical strategies they could begin implementing immediately.
Jo Peterson
CIO, Clarify360
Where Has Mark Spoken?
According to venue records, Mark has delivered keynotes at: RSA Conference · Oracle CloudWorld · Cisco Partner Summit · Dell Technologies World · IBM Think · T-Mobile Events · Gartner Security & Risk · InfoSecurity · ISACA Conferences · ISSA Events · Cloud Security Alliance · CyberSecurity Summit · BSides · FLGISA · MISAC · SMU Cox School of Business · and 100+ more.
How Do You Book Mark Lynd for Your Event?
The booking process is straightforward and typically completes within 3 business days. Mark customizes every keynote to the audience, industry, and event objectives.
- Submit an inquiry. Fill out the contact form with your event date, audience, and objectives. Response within 48 hours.
- 15-minute discovery call. Discuss your event in detail, including audience makeup, key messages, and desired outcomes.
- Proposal & contract. Receive a tailored proposal with format options (keynote, workshop, panel), fee, and travel terms.
- Customization. Mark customizes content to your audience, industry examples, and desired takeaways.
- Expert delivery. Mark brings 25+ years of real-world executive experience to every stage.
Frequently Asked Questions
What makes Mark Lynd uniquely qualified as an AI infrastructure keynote speaker?
What does Mark cover at an AI infrastructure keynote?
Who is the right audience for this keynote?
Does Mark cover the AI sustainability and water conversation?
Does Mark cover sovereign AI and regional cloud?
Does Mark address AI security inside the AI infrastructure conversation?
Is this keynote a fit for board and audit-committee audiences?
Can Mark customize for our specific industry?
What size audiences does Mark speak to?
Does Mark deliver virtual and hybrid AI infrastructure keynotes?
How much does it cost to book Mark for an AI infrastructure keynote?
What is the booking and prep process?
Ready to Book Mark?
Mark customizes every keynote to the audience, industry, and event objectives. Contact the booking team to check availability.
100+ keynotes · audiences from 50 to 5,000+
Formats: In-person, virtual, hybrid · international available
Duration: 30–120 minutes depending on format
Custom by event · educational pricing available