Skip to content
The tech talent gap refers to the shortage of skilled IT professionals needed to support modern infrastructure and digital transformation.

The technology hiring landscape has fundamentally changed in the past few years. Following the pandemic reopening, tech job postings more than doubled as remote work exploded, venture capital flooded the sector, and low interest rates fueled aggressive growth. Yet, when the Federal Reserve raised interest rates in 2022, tech job listings plummeted. The collapse continued through 2023 and then stabilized at significantly lower levels.

Now, tech job postings sit roughly 35% below pre-pandemic levels and have stayed relatively steady since mid-2025. Even with widespread layoffs and economic uncertainty, companies still struggle to attract and retain the technical talent they need to grow. The challenge isn’t volume since recruiting teams are usually flooded with applications. The market itself has shifted. While there’s a surplus of candidates for generalist technical roles, deeply specialized positions remain extremely difficult to fill.

For talent acquisition leaders, the pressure is mounting from multiple directions. Hiring managers want specific skill sets, finance wants lower costs, and leadership wants faster time-to-hire. One thing is clear: Success isn’t about offering the highest salaries. It’s about rethinking your entire approach to finding, evaluating, and keeping technical talent.

Here are the strategies and practices that are actually producing results in today’s technical hiring market.

Stop Hunting for Unicorns

The unicorn hunt needs to stop. Too many organizations create job descriptions that read like impossible checklists: seven years of experience with a framework that’s only existed for four, mastery of twelve different programming languages, expert-level knowledge in cloud architecture, DevOps, security, and data science. These descriptions tend to screen out anyone who doesn’t tick every box, even when most of those skills are transferable or learnable in weeks.

This approach ignores how fast technology changes. Today’s must-have stack could be outdated in 18 months. More importantly, it prices you out of the market. When you’re looking for candidates with every possible qualification, you have to pay premium rates. Then you’re stuck choosing between people you can’t afford and roles that stay open for months.

The better approach? Work with the hiring manager to identify three non-negotiables and evaluate everything else based on adaptability and learning ability. A strong engineer who knows how to learn will outperform someone with the perfect resume but no curiosity.

What Actually Attracts Technical Talent

Candidates need to understand their “why.” Technical talent wants to know why this role matters to them personally. What problem will they solve? How does their work connect to the bigger picture? What impact will they have if they succeed? If hiring managers can’t articulate this clearly during the interview process, they’re already losing people.

Demonstrate autonomy and trust early. Top performers want the freedom to solve problems their way, not follow a prescribed playbook. They want to put their fingerprints on something and make it uniquely theirs. That means giving them ownership over projects, not just tasks. The interview process itself should reflect this culture. Talk about how decisions get made, how much independence technical employees have, and what kind of trust they’ll be given from day one.

Make career growth concrete and visible. Technical candidates want to know what certifications and training you’ll support, what skills they can develop, and what their trajectory could look like in the first year. Vague promises about “growth opportunities” don’t resonate with people who are evaluating multiple offers. Be specific about what advancement actually looks like in your organization and show them there’s a clear path forward, not just a job.

Making the Final Decision: What Tips the Scale

Executive engagement makes a bigger difference than most companies realize. For a recent VP-level hire, the candidate had multiple offers on the table, including ones that paid more. What made the difference? Direct conversations with the CEO about vision and strategy, a meeting with the CFO to discuss equity and long-term value, and time with peers to understand the team dynamics and day-to-day reality of the role.

Flexibility on the logistics mattered as well. In this case, being willing to accommodate specific needs during the transition showed good faith. It’s not about bending over backward for every candidate but demonstrating that you’re willing to work with people when it matters.

Why Technical Employees Stay (or Leave)

Career growth opportunities and autonomy drive retention more than anything else. Technical employees want to see a clear path forward, whether that’s deepening expertise, taking on leadership responsibilities, or moving into new technical domains. They also want the freedom to make something uniquely theirs. Ownership over projects and outcomes matters far more than perks or ping pong tables.

One of the most underrated retention factors is cross-organizational collaboration. When employees feel siloed or stuck working only within their immediate team, they start looking elsewhere. Technical talent stays engaged when they can reach out to anyone across the organization for support, regardless of department or hierarchy. That kind of collaborative environment is evidence of a culture where problem-solving matters more than politics.

For our company, data center tours can be very effective, too. Seeing the actual facilities, understanding the scale of operations, and grasping how their role would impact real infrastructure makes the position more tangible for prospective employees. In these cases, candidates can visualize their contribution in a way that phone screens and slide decks never could.

Navigating AI in the Hiring Process

AI has changed resume screening and interviewing in ways that create new challenges for recruiters. Resumes now include obvious AI-generated content that doesn’t match the candidate’s actual experience. Interview red flags include unusual pauses, backgrounds that seem to shift or flicker, and responses that feel too polished or generic. When candidates are using AI assistance during live interviews, it becomes impossible to assess their real capabilities.

The human element remains essential despite these technological shifts. While AI tools can help with initial resume screening and ranking candidates against job requirements, someone still needs to read each application with an eye for potential. A candidate might not match the current role perfectly but could be ideal for another open position, and AI consistently misses those connections.

Authenticity has become non-negotiable in the hiring process. Clear evidence that a candidate is using AI during interviews or has fabricated his or her resume with AI-generated content leads to immediate disqualification. The focus has shifted to verifying that the person you’re interviewing is the same person who will show up on day one.

Getting Technical Hiring Right

If I had one piece of advice for organizations trying to attract and hire top technical talent, it would be this: The “why” discussed earlier needs to extend beyond the interview process. Create it for every role and make sure the candidate fully understands it before he or she accepts an offer. Help candidates see their piece of the bigger picture. Walk them through what the first 30, 60, and 90 days actually look like. Explain how they can grow within the company and what success means in concrete terms.

Too many candidates accept offers without truly understanding the daily reality of the work or how their contribution matters. When that clarity exists from day one, you get better hires who stay longer and perform at a higher level. Everything else flows from that foundation.

Data center construction trends reflect changes in power density, cooling design, sustainability requirements, and deployment speed.

The data center construction industry enters 2026 facing challenges that would have seemed unimaginable just two years ago. Rapid growth is straining every aspect of traditional construction approaches, from securing skilled labor to managing unprecedented project scales. 

The fundamentals that worked for decades no longer apply. Single campuses now require 4,000 workers instead of 750. Customer specifications shift mid-deployment as technology evolves faster than buildings can rise. Construction teams must adapt to a new reality where flexibility and forward planning matter more than following established playbooks. 

Based on what we’re seeing across our projects, here are the trends that will define data center construction in 2026. 

Talent Shortage Will Be the Biggest Challenge 

The skilled labor shortage in the data center industry will intensify significantly in 2026 as leased capacity from 2024 and 2025 finally comes online. The second half of 2026 into 2027 will see massive activation across the country, and the industry simply doesn’t have enough qualified workers to meet demand. 

Project scales have multiplied beyond recognition. Where peak crew sizes once reached 750 workers, sites like DataBank’s Red Oak campus will hit 4,000 to 5,000 workers by early 2026. That’s the size of a small city, requiring entirely different management approaches than data center builds in the past. 

Labor costs are rising accordingly. Workers are relocating from markets like Arizona, where power constraints have slowed construction, to booming regions like Dallas. Increased wages, per diem expenses, and relocation costs are adding up. Some operators are even establishing on-site housing with high-end amenities to attract workers. 

The operators who secure skilled labor now, through early contractor commitments and creative workforce strategies, will have significant advantages over those waiting until projects break ground. 

Costs Will Continue Rising 

Cost increases are hitting every aspect of data center construction: power, materials, and especially people. While DataBank maintains multi-year stocking programs to lock in material pricing and minimize tariff impacts, labor cost escalation is harder to control given the talent shortage. 

The type of construction matters significantly for cost management. Operators building exclusively for AI workloads can opt for cheaper approaches like steel buildings with metal panels rather than robust precast structures. However, versatile construction capable of serving any client type costs more upfront while providing long-term flexibility. 

Leasing rates are rising to reflect these realities. The question becomes whether operators can absorb increased costs or must pass them to customers. Those with strong vendor relationships, economies of scale through project bundling, and long-term procurement strategies will be better positioned to manage cost pressure than competitors making reactive, project-by-project decisions. 

Construction Innovation Will Address Labor Challenges 

The industry is turning to technology and new construction methods to offset the skilled labor shortage. AI tools are already analyzing pull schedules to identify float and delays that would take humans hours to find manually. Robotics like autonomous site scanners perform 360-degree analysis, comparing progress against 3D BIM models with greater accuracy than human inspections. 

These tools won’t immediately replace construction roles. The industry still needs experienced boots on the ground, particularly the skilled general foremen and project leaders who manage complex deployments. Quality control and administrative positions may evolve over the next five years as AI handles more analytical work. 

Modular construction and prefabrication are accelerating in parallel. Electrical components, power rooms, and building modules assembled in controlled factory environments address labor constraints more effectively than on-site work. A crew of 20 to 50 workers in a factory can be more productive than managing thousands on a sprawling campus, where workers lose two hours daily just commuting to and from the site. 

Managing 4,000-plus workers on a single site can also lead to safety challenges that didn’t exist at smaller scales. Setting clear rules and maintaining safety protocols becomes critical when construction sites rival small cities in population. 

Financial Pressure Will Drive Market Consolidation 

Power infrastructure costs are creating a significant barrier to entry. Operators pursuing geothermal energy, high-pressure gas lines with turbines, or small modular reactors face capital requirements in the hundreds of millions before a single tenant moves in. Without committed customers, it’s a major challenge to secure approval for an additional $200 million to $300 million in power infrastructure on top of base construction costs. 

Utility commitments amplify the risk. Non-refundable deposits of multi-millions of dollars force developers to decide who will roll the dice and who won’t. Operators with purchase power agreements and facility engineering agreements already locked in can build more speculatively because they’ve de-risked the power equation. Those without these commitments will struggle. 

Creditworthiness issues compound the challenge. Many AI-focused tenants lack strong credit profiles, forcing hyperscale cloud providers to serve as backstops for lease financing. This dynamic will intensify in 2026, particularly for shorter-term leases. The combination of massive power infrastructure costs and tenant credit concerns will separate well-financed operators with proven execution from speculative players who entered the market recently. Some will exit entirely when the financial reality becomes insurmountable. 

Customer Design Requirements Will Keep Changing 

Every liquid cooling deployment involves different customer requirements, and specifications continue shifting mid-project as technology evolves. For example, at our Red Oak campus, a customer’s requirements shifted from nine liquid-cooled data halls to seven mid-deployment as network capacity needs outweighed GPU processing requirements. 

Design variations are significant. Some deployments require complex programmable logic controller integration while others use simpler approaches. Cooling distribution unit placement varies from exterior-mounted to in-cabinet configurations. Air cooling remained relevant even in late 2025, with enterprise customers and some financial firms still specifying significant air-cooled capacity alongside liquid infrastructure.  

The key is building flexibility into designs from the start rather than attempting expensive retrofits later. 

Permitting and Community Engagement Will Intensify 

Despite administration comments about fast-tracking AI data centers, the reality on the ground is moving in the opposite direction. Permitting processes are becoming more delayed, with additional third-party engineers reviewing applications and more stringent approval requirements. This trend will continue as a significant challenge over the next few years. 

Community engagement has become essential to successful project development. Many communities harbor misconceptions about data centers, conflating them with AI development or robotics manufacturing rather than understanding they enable everyday digital services. Education and early relationship building with local stakeholders are now critical components of the development timeline, not optional add-ons. 

Building for What’s Next 

The only certainty in data center construction for 2026 is uncertainty. Technology will evolve faster than project timelines. Customer requirements will shift mid-deployment. Costs will surprise even the most careful planners.  

Success belongs to construction teams built for adaptation rather than those optimized for efficiency in stable conditions. The ability to pivot quickly, maintain optionality in designs, and solve problems creatively matters more than following the playbook that worked last decade. 

Want to see how DataBank is navigating the evolving data center construction landscape? Check out DataBank Digest for more insights on managing talent challenges, cost pressures, and the innovative approaches shaping infrastructure development in 2026.

Corporate social impact programs highlight how organizations support communities, education, and social initiatives.

At DataBank, we believe that being a good neighbor means more than just operating reliable infrastructure. It means showing up for the communities where our teams live and work, supporting the people who need it most, and creating a culture where giving back is woven into everything we do.

DataBank Cares program logo

DataBank Cares empowers employees to give back through volunteer time, charitable donations, and community partnerships across all our locations.

2025 was a remarkable year for DataBank Cares, our company-wide initiative that empowers employees to lead community engagement efforts through volunteer time, organized campaigns, and direct support. From coast to coast, DataBank teams demonstrated what’s possible when technology companies commit to making a tangible difference beyond the data center walls.

The numbers tell an impressive story: 857 hours of Volunteer Time Off (VTO) contributed across dozens of initiatives, thousands of pounds of food and supplies donated, hundreds of letters written to troops overseas, and a groundbreaking employee hardship fund that raised over $40,000.

Yet, behind every statistic are the people who made it happen: employees who chose to give their time, energy, and resources to support their neighbors.

Here’s how DataBank teams showed up in 2025.

Environmental Stewardship: Earth Day and Beyond

We kicked off the year in April with Earth Day initiatives that demonstrated our commitment to environmental responsibility in the communities we serve. Teams across our network organized cleanup events and donation drives that made immediate, visible impact.

In Denver, our metro team joined forces with 250 volunteers in partnership with Protect Our Rivers, removing nearly 5,000 pounds of trash from local waterways. The effort represented hours of hands-on environmental work that directly improved the natural spaces on which our community depends.

DataBank Earth Day volunteers

DataBank employees joined 250 volunteers in partnership with Protect Our Rivers to remove nearly 5,000 pounds of trash from local waterways during Earth Day 2025.

Meanwhile, at our Salt Lake City campus, the team organized a comprehensive donation drive for the SLC Rescue Women’s Shelter, collecting 24 bags and 7 boxes of clothing and feminine products, along with $400 in cleaning supplies. The initiative addressed immediate needs while demonstrating how data center teams can mobilize resources to support vulnerable populations.

Kindness in Action: Supporting Women Experiencing Homelessness

In May, 143 DataBankers came together—virtually and in person—for one of our most ambitious initiatives of the year. Led by our Women of DataBank employee resource group, teams assembled over 650 Kindness Kits for women and people experiencing homelessness across Denver, Houston, Dallas, and Salt Lake City.

These weren’t just supply kits. They were carefully curated packages of essential items that provide dignity and support during difficult circumstances. The campaign brought together employees from multiple locations who shared a commitment to addressing homelessness with compassion and practical assistance.

Women of DataBank volunteers

Women of DataBank employee resource group members assemble Kindness Kits during a company-wide initiative that created over 650 kits for women experiencing homelessness across four cities.

The initiative reflected DataBank’s belief that employee resource groups aren’t just internal networks. They’re engines for community impact. When employees lead efforts they’re passionate about, the results are more authentic, more meaningful, and ultimately more effective.

Fighting Food Insecurity Across Four Markets

From August through September, DataBank launched a company-wide food drive that mobilized teams across our 65+ data center locations. The results exceeded our own expectations:

Combined, these efforts delivered nearly 3,000 pounds of food and essential supplies to communities across the country. You can read more about DataBank’s food drive success here.

Honoring Those Who Serve

In one of our most personally meaningful initiatives, DataBank teams across eight data centers wrote more than 500 letters of gratitude to U.S. military service members serving at home and abroad. Teams in Atlanta, Denver, Dallas-Fort Worth, Newark, Houston, New York, Orangeburg, and Salt Lake City participated, with 78 DataBankers from 10 data centers contributing letters sent through “A Million Thanks.”

DataBank Letters to the Troops

DataBank employees across eight locations wrote over 500 letters of gratitude to U.S. military service members through A Million Thanks.

Research consistently shows that letters from home significantly boost morale and mental wellbeing for deployed troops. These personal connections remind service members that their sacrifices are recognized and valued by the communities they protect.

The campaign also highlighted individual dedication, like Senior Security Architect Stephen Maiorca, who has been handcrafting wooden pens for deployed service members since 2017. Each pen requires about two hours of work, and Maiorca donates them to “For The Troops,” a Simi Valley organization that creates care packages. His ongoing commitment, inspired by veterans in his own family, exemplifies the personal passion many DataBank employees bring to community service.

Investing in Education and Future Opportunities

DataBank’s partnership with Red Oak Public Library in Texas brought hands-on technology directly into students’ hands. Through DataBank’s sponsorship, the library purchased five 3D printers that students can check out and use at home.

During the launch event for the library’s STEAM (Science, Technology, Engineering, Arts, and Mathematics) class, students gathered to watch the technology in action, selecting from pre-loaded designs and learning how these machines transform digital files into physical objects. Each student left with a 3D-printed holiday figurine he or she painted and personalized. Yet, more importantly, students left with deeper understanding of modern manufacturing technology and the creative possibilities it opens up.

DataBank Red Oak community partnership

DataBank partners with Red Oak Library to bring hands-on STEAM learning and 3D printing technology to local students.

The take-home aspect makes this program is particularly impactful. By allowing families to check out the printer kits, the library removes barriers to technology access and gives students the chance to experiment at their own pace. When you’re building significant infrastructure like our Red Oak campus—a 292-acre development that will deliver eight data centers and 3.4 million square feet of AI-ready infrastructure— being a good neighbor means investing in the young people who represent the future workforce.

Holiday Season Giving

DataBank teams closed out the year with widespread holiday initiatives. Across eight data centers, employees donated over 500 toys and $200 to the U.S. Marine Corps Toys for Tots drive, spreading holiday joy to children in need.

Throughout the year, DataBankers also participated in the #DataCentersSleevesUp2025 virtual blood drive and helped build homes for the Sleep in Heavenly Peace Foundation, demonstrating the breadth of causes our teams support.

Supporting Each Other: The DataBank Cares Fund

Perhaps, the most significant development in 2025 was the launch of the DataBank Cares Fund, created to help employees facing unforeseen disasters or personal hardships. The fund relies on support from DataBank Holdings Ltd. and individual donations from employees, partners, and the general public.

Launched in November 2025, the employee giving campaign exceeded all expectations. DataBank employees gave anonymously and surpassed the company’s internal goal, raising nearly $23,000. DataBank matched these donations, creating a fund of over $40,000 to assist any DataBank employee in need.

This initiative reflects a fundamental truth: When we take care of each other, we build stronger teams and stronger communities. Every donation contributes to a safety net that provides grants when they’re needed most.

A Culture of Giving Back

The success of DataBank Cares in 2025 demonstrates what happens when employees lead initiatives about which they’re passionate. From employee resource groups organizing Kindness Kits to individuals handcrafting gifts for troops, these efforts reflect genuine commitment rather than corporate obligation.

Operating critical infrastructure comes with responsibility to support the communities where we’re located. DataBank teams find meaningful ways to give back year-round—from food banks and schools to children’s hospitals and our own employee support systems.

We’re proud of what our teams accomplished in 2025 and grateful to the organizations serving people in need. As we look toward 2026, we remain committed to being the neighbors our communities deserve—present, engaged, and ready to help when it matters most.

Together, we’re building stronger communities, one act of service at a time.

Want to see more stories about DataBank’s community involvement and industry leadership? Check out DataBank Digest for more insights on data center industry trends and infrastructure strategy.

AI-powered network operations use machine learning and automation to monitor, optimize, and manage network performance at scale.

Today, enterprise AI deployments are reshaping network demands faster than traditional infrastructure management can adapt.

Consider the scale: Machine learning models require massive datasets to move quickly between storage, compute, and processing nodes, often saturating network links for hours or days during training runs. Additionally, inference workloads add another layer of complexity, demanding microsecond-level latency to deliver real-time results.

As organizations scale from pilot projects to production AI, these demands multiply across hundreds or thousands of interconnected systems, creating network complexity that manual management simply can’t handle.

When Manual Management Hits Its Limits

For data center providers, this creates a difficult equation. Customers expect their AI workloads to run at peak performance with minimal latency and zero downtime. However, delivering that level of service means managing exponentially more network traffic, monitoring thousands of potential failure points, and responding to issues before they can affect workloads.

Yet traditional network operations approaches such as reactive monitoring, manual optimization, and scheduled maintenance windows simply can’t scale to meet these demands without adding substantial cost and operational overhead.

The answer lies in applying AI to the very infrastructure that supports AI workloads. By using machine learning to monitor network health, predict bottlenecks, and automatically adjust configurations, data center operators can deliver the performance AI customers require while maintaining operational efficiency. This approach rests on three fundamental capabilities: observability, optimization, and automation.

The Three Pillars of AI-Powered Network Operations

These three pillars work together to transform how data centers manage connectivity infrastructure. Observability provides deep visibility into network performance and behavior. Optimization uses that visibility to continuously improve how the network operates. Automation reduces manual intervention while enabling faster, more consistent responses to changing conditions. Together, they allow operators to maintain the performance and reliability AI applications require at scale.

Here’s a closer look at how each capability works in practice.

Observability

Traditional network monitoring provides basic metrics like bandwidth utilization and uptime percentages. Yet AI workloads create traffic patterns that shift dynamically based on training schedules, data pipeline flows, and inference request volumes. Understanding network health requires deep visibility into these patterns.

AI-powered observability platforms process massive amounts of network telemetry data in real time, detecting anomalies before they become problems. Machine learning models can detect subtle patterns that indicate a switch is beginning to fail, a fiber connection is degrading, or traffic is routing suboptimally. This level of insight allows operators to understand not just what is happening on the network, but why it’s happening.

The result is a shift from reactive troubleshooting to proactive management. Instead of waiting for customers to report performance issues, operators can see problems developing and address them before they impact workloads.

This visibility becomes especially critical as AI deployments scale and network complexity grows beyond what manual monitoring can effectively track. Strategic interconnection decisions that reduce latency and improve path diversity make this observability even more valuable, allowing operators to optimize performance across increasingly complex network topologies.

Optimization

Network optimization has traditionally relied on periodic analysis and manual configuration changes. Engineers review performance data, identify bottlenecks, and adjust routing or bandwidth allocation during maintenance windows. This approach works for relatively stable workloads but struggles with the dynamic demands of AI infrastructure.

AI-powered optimization operates continuously, analyzing traffic patterns and automatically adjusting network configurations to improve performance. Machine learning algorithms can predict when certain network paths will become congested and proactively reroute traffic. The right connectivity architecture makes this optimization possible by providing the diverse paths and direct connections AI workloads require. They can allocate bandwidth dynamically based on workload priority, ensuring critical AI training jobs get the resources they need while maintaining service levels for other applications.

This continuous optimization delivers measurable improvements in network efficiency and application performance. Latency drops, throughput increases, and resource utilization improves without requiring constant manual intervention.

Automation

Network engineers spend significant time on routine tasks like responding to alerts, updating configurations, and troubleshooting connectivity issues. As network complexity grows with AI workload demands, these manual processes become bottlenecks that limit how quickly operators can respond to problems or scale infrastructure.

AI-powered automation handles many of these routine tasks without human intervention. Systems can automatically remediate common issues, adjust configurations based on changing conditions, and even predict maintenance needs before equipment fails. This reduces the operational burden on engineering teams while improving response times from hours or days to minutes or seconds.

Automation also brings consistency to network operations. Configurations are applied uniformly across infrastructure, reducing the risk of human error. Engineers can focus on strategic initiatives rather than firefighting daily issues, improving both operational efficiency and service quality.

Together, these three capabilities create a network management approach that matches the scale and complexity of modern AI workloads. Observability reveals what’s happening across the infrastructure, optimization ensures it operates efficiently, and automation enables rapid response without manual intervention.

Infrastructure Ready for AI’s Future

As AI workloads continue to grow in scale and complexity, the network infrastructure supporting them must evolve beyond traditional management approaches. AI-powered network operations provide the observability, optimization, and automation capabilities needed to deliver the performance and reliability enterprise AI applications demand.

By applying intelligence to the network layer itself, data center operators can meet customer expectations while maintaining operational efficiency at scale. The result is infrastructure that doesn’t just support AI workloads but adapts and improves alongside them.

Want to stay ahead of infrastructure technology trends like these? Check out DataBank Digest for more insights on data center industry trends and infrastructure strategy.

Corporate community programs support education and workforce development by providing resources, technology access, and learning opportunities.

When you think about what makes a community thrive, technology access is often at the top of the list. That’s why we’re excited to share that DataBank recently partnered with the Red Oak Public Library in Texas to help bring hands-on 3D printing technology directly into the hands of young students.

Through DataBank’s sponsorship, the library was able to purchase five 3D printers that students can check out and use at home. During the recent launch event for the library’s STEAM (science, technology, engineering, arts, and math) class, students gathered around the printers to watch the technology in action, selecting from pre-loaded design options and learning exactly how these machines transform spools of filament into finished objects.

Students learning 3D printing technology

DataBank’s partnership with Red Oak Library brings cutting-edge STEAM resources and hands-on learning to local students.

Each student left with a 3D-printed holiday figurine (a snowman or gnome) that they painted and personalized. Even better, they left with a deeper understanding of how modern manufacturing technology works and the creative possibilities it opens up.

Taking Technology Home

The take-home aspect of this program is particularly meaningful. By allowing families to check out the 3D printer kits, the library is removing barriers to technology access and giving students the chance to experiment and create at their own pace.

“We’re thrilled to offer these take-home 3D printers to our community,” said Heather Fuller, Red Oak Public Library Director. “When students can take this technology home and experiment at their own pace, it opens up possibilities that go far beyond a single class session. They’re not just learning how the technology works, they’re discovering what they can create with it.”

Building Community Through Infrastructure

This partnership reflects DataBank’s commitment to supporting the communities where we operate. Red Oak is home to our newest data center campus—a 292-acre mega-development that will deliver eight data centers and 3.4 million square feet of AI-ready infrastructure.

When you’re building something this significant in a community, being a good neighbor means more than just construction jobs and tax revenue. It means investing in the people who live here, especially young students who represent the future workforce. Programs like this 3D printing initiative give students hands-on access to technology that builds skills and sparks curiosity—the same innovative thinking that drives the work happening at our campus. When students in our communities have access to tools that open doors to technical careers, everyone benefits.

We’re proud to support the Red Oak Library’s STEAM programming and look forward to seeing what these young innovators create.

Data center security predictions highlight emerging risks, regulatory pressures, and evolving threat models that influence infrastructure protection strategies.

As we enter 2026, security considerations for data center operators continue to evolve in ways that demand both vigilance and adaptability. While I’m cautiously optimistic that we’ll see some geopolitical tensions ease by year’s end, with conflicts in Ukraine and the Middle East potentially moving toward de-escalation, the threat environment isn’t simplifying. It’s shifting.

Nation-state actors are getting more sophisticated. Insider threats are multiplying. The attack surface keeps expanding as AI workloads drive unprecedented connectivity requirements. Regulatory frameworks are tightening globally, creating new compliance burdens on top of operational security challenges.

Here’s what’s keeping me up at night, and what should be on your radar.

The Insider Threat Goes Nation-State

The most concerning physical security trend we’re tracking is the emergence of sophisticated insider threats tied to nation-state actors.

Industry intelligence suggests a troubling possibility: Individuals connected to foreign governments may be presenting themselves as legitimate job candidates and attempting to get hired at data centers, in technology roles in general, or in situations where intellectual property has value, to position themselves for future directed activities.

In some cases, reports indicate intellectual property theft. In others, these individuals may be lying dormant, potentially waiting to disrupt operations by pulling switches or disconnecting critical infrastructure if geopolitical tensions escalate. Security professionals are increasingly warned about this threat vector, and it’s fundamentally changing how we need to think about personnel security and background verification.

This isn’t paranoia. This is a documented threat pattern that requires data center operators to strengthen pre-employment screening, maintain heightened awareness of unusual employee behavior, and implement security controls that assume insider risk rather than blindly trusting credentials.

Biometric Security Faces Its Deepfake Moment

Physical security systems have increasingly relied on biometric authentication such as facial recognition and fingerprint scanners as more reliable than or as an addition to traditional access controls. That reliability is now under assault.

We’re seeing rapid advances in biometric spoofing, where attackers use 3D-printed face models, latex fingerprints, and other techniques to defeat biometric systems. What makes this particularly concerning is the convergence with AI. Threat actors can now take detailed 3D scans and use AI to generate remarkably convincing physical models that can fool facial recognition systems. What used to be exciting material for movies is now reality.

The fingerprint spoofing is slightly less advanced, but attackers have successfully created latex glove overlays that mimic fingerprints well enough to bypass some scanning systems. As these techniques become more accessible and refined throughout 2026, data center operators will need to implement multi-factor authentication even for biometric systems and consider liveness detection technologies that can distinguish between real human features and sophisticated fakes.

AI Becomes Both Weapon and Battleground

The AI security threat is no longer theoretical. We’ve already seen instances in late 2025 where AI systems were weaponized to launch attacks against multiple organizations. One AI platform was exploited to target approximately 30 organizations before the provider detected and stopped the attack. Unfortunately, this is just the beginning.

In 2026, we’ll see AI evolve as both an attack tool and an attack surface. Smaller AI providers without robust security controls will become particularly vulnerable, and their tools could be hijacked to conduct sophisticated attacks. More concerning is the emergence of what security researchers call “ghost inputs.” These are carefully crafted prompts designed to trick AI systems into performing malicious actions while believing they’re doing something legitimate.

The intersection of AI and physical security creates additional vulnerabilities. Attackers are already exploring how to use AI to generate the biometric spoofing tools I mentioned earlier, creating a feedback loop where AI enables more sophisticated physical intrusions.

Data center operators should be evaluating their AI security posture now, implementing strong governance frameworks, and recognizing that AI systems need the same security scrutiny as any other critical infrastructure component.

Criminal Campaigns Fill the Geopolitical Void

Here’s a prediction that might surprise you: as nation-state cyber warfare potentially de-escalates with improving geopolitical conditions, we’re going to see criminal advanced persistent threats (APTs) surge to fill that void.

These aren’t your typical ransomware crews. These are sophisticated organizations, often with plausible deniability connections to nation-states, focused on economic espionage, high-value intellectual property theft, and supply chain extortion. They’ve realized that disrupting supply chains creates cascading economic impacts that can be monetized multiple ways.

Our just-in-time supply chain infrastructure makes us particularly vulnerable. When critical systems get disrupted—whether it’s airline operations, shipping logistics, or manufacturing coordination—the financial impact ripples through entire economic sectors. These criminal APTs are betting that companies will pay to avoid or stop these disruptions.

For data center operators, this means renewed focus on supply chain security, vendor risk management, and ensuring that your customers’ critical operations remain resilient against these sophisticated threats.

The Persistent Danger of Social Engineering

Despite all our technological advances, business email compromise and social engineering remain devastatingly effective. We’re still seeing incidents where employees click malicious links despite extensive training coupled with advanced detection tools.

Deepfake technology is making social engineering even more dangerous. We’ve seen cases where attackers impersonate vendors with remarkable precision, resulting in millions of dollars redirected to fraudulent accounts before detection. As deepfake audio and video become more convincing throughout 2026, the verification challenge becomes exponentially harder.

Our response includes investing in AI-powered email security solutions—using safe, known AI technologies to detect and prevent compromise attempts—and implementing stronger verification protocols for any financial transactions or sensitive communications.

A Growing Concern: Community Backlash

Later in 2026 and beyond, I’m watching another threat emerge from an unexpected direction: community opposition to data centers potentially escalating beyond political channels into physical risk.

We’re already seeing significant citizenry backlash at zoning meetings, with concerns about power consumption, water usage, and community impact. Many of these concerns are based on incomplete or incorrect information. Right now, this remains in the realm of discussion. However, I am concerned about narratives that could radicalize opposition—stories suggesting that data centers have made housing unaffordable, driven up utility costs, or forced families from generational property.

If someone becomes convinced that a data center has personally harmed him or her economically, that opposition could, potentially, manifest as physical threat. This is still a long-range concern, but it’s something we’re monitoring carefully, and it underscores the importance of proactive community engagement and transparent communication about data center operations.

Getting Security Right in 2026

The security landscape for 2026 demands that we stay ahead of evolving threats without becoming paralyzed by them. At DataBank, we’re addressing these challenges through updated physical security technologies, enhanced detection capabilities, and stronger protocols around both cyber and physical access.

Here’s my advice to the industry: Don’t over-fortify to the point where you create more problems than you solve. When data centers start looking like Fort Knox with robo-dogs and drones, you’re not just deterring threats; you’re drawing attention and fueling the very myths and concerns that create community opposition.

Smart security is about layered, proportionate defenses that protect assets without broadcasting paranoia. It’s about recognizing that in 2026, security threats span from nation-state actors to AI vulnerabilities to supply chain risks, and addressing them all requires both sophistication and restraint.

The year ahead will test our ability to adapt to new threat vectors while maintaining the trust and confidence of the communities we serve. That’s a challenge I’m confident our industry can meet.

Cloud repatriation involves moving workloads from public cloud platforms back to private or colocation infrastructure due to cost, performance, or compliance concerns.

This is Part 2 of our series on cloud repatriation. Read Part 1 to learn why 86% of CIOs are rethinking their cloud strategy.

DataBank recently published a white paper examining the reality of cloud repatriation, exploring why enterprises are moving workloads back from public cloud, the hidden challenges of repatriation itself, and strategies for avoiding these pitfalls altogether. To learn more, download “The Reality of Cloud Repatriation” now.

Understanding why organizations want to leave the cloud represents only half the story. Many IT leaders recognize the cost overruns, compliance complexities, and support limitations driving infrastructure reconsideration. They see the business case for repatriation clearly. Yet what catches them off guard are the obstacles that emerge when they actually try to execute the migration.

Organizations often approach repatriation expecting a straightforward reversal of their cloud migration, only to discover a new landscape of technical, operational, and financial barriers. The very features that made cloud attractive in the first place can become handcuffs when trying to leave.

Understanding these challenges upfront helps organizations make realistic assessments of repatriation costs and complexity, or better yet, avoid the need for repatriation altogether through smarter initial decisions.

The Exit Fee Surprise

The most immediate shock in cloud repatriation often comes from data egress charges. Cloud providers design their pricing to encourage adoption. Data ingress is typically free or minimal, making it easy and inexpensive to move workloads into their platforms. Data egress, however, carries substantial fees that many organizations have already experienced during regular operations like analytics processing or backups.

When repatriation requires moving terabytes or petabytes of data out permanently, these charges can represent a significant barrier to exit. Organizations that migrated years of historical data, customer records, transactional information, and operational logs into cloud storage discover that extracting it all comes with a price tag that can reach tens or hundreds of thousands of dollars, depending on volume.

These egress fees are often the first domino to drop. Application downtime during migration translates to lost revenue or productivity. Staff time dedicated to planning and executing the migration diverts resources from other strategic initiatives. Performance testing and validation require dedicated environments and personnel. The risk of business disruption if transitions don’t go smoothly introduces both direct costs and reputational concerns that executives must weigh carefully.

For mission-critical applications, organizations often need to maintain parallel environments during transition periods, effectively paying for both cloud and alternative infrastructure simultaneously. For example, a financial services company cannot simply shut down its cloud-based trading platform and hope the new infrastructure works perfectly.

They must run both systems in parallel, validate performance under real-world conditions, and gradually shift traffic only after proving the new environment meets their requirements. This parallel operation extends both the duration and expense of repatriation significantly.

The timeline itself becomes a cost factor. Unlike faster cloud migrations that often use lift-and-shift approaches, repatriation requires careful execution to avoid service disruptions. What organizations hoped might take weeks can stretch into months, with associated costs accumulating throughout the entire transition period.

From Elastic to Predictive

One of the most significant operational shifts during repatriation involves moving from cloud’s elastic scaling model to capacity planning for private environments. In public cloud, infrastructure automatically responds to demand spikes or scales down during quiet periods. Repatriation requires a fundamentally different mindset where IT teams must determine appropriate infrastructure sizing upfront, forecast growth requirements, and maintain sufficient capacity buffers without wastefully overprovisioning.

The risks of getting this wrong cut both ways. Under-provisioning leads to performance problems and potential business disruptions. When an e-commerce platform cannot handle seasonal traffic spikes because infrastructure capacity was sized too conservatively, revenue suffers directly. Overprovisioning wastes capital on unused capacity that sits idle. Finding the right balance requires experience and analytical capabilities that organizations may need to rebuild or acquire.

Performance expectations also require recalibration during repatriation. While properly designed private infrastructure often delivers better and more consistent performance than public cloud, the transition period can introduce challenges as teams optimize configurations and tune systems for specific workloads.

The Reality Check

Cloud repatriation is not a simple undo button for cloud adoption. The process introduces substantial technical, operational, and financial challenges that organizations must navigate carefully. Success requires realistic planning, adequate resources, and often significant system modifications.

These challenges explain why prevention proves far more effective than attempting to unwind poor cloud strategies after the fact. Organizations that partner with infrastructure providers offering multiple deployment options, maintain vendor agnosticism from the outset, and make strategic workload placement decisions based on actual business requirements can adapt as needs evolve without facing costly repatriation obstacles.

Our white paper explores these repatriation challenges in greater depth and provides concrete strategies for avoiding them through smarter initial infrastructure decisions. Download “The Reality of Cloud Repatriation“ to learn how strategic infrastructure planning and flexible deployment models prove far more effective than reactive repatriation.

In our final post in this series, we’ll examine prevention strategies that help organizations avoid repatriation challenges entirely through flexible infrastructure planning and strategic workload placement.

Data centers drive economic impact through infrastructure investment, job creation, and support for digital services.

A version of this article first appeared in the Culpeper-Star Exponent. 

Communities looking to attract business and technology investment face a common dilemma: how to drive economic growth while addressing concerns related to sprawl, the project’s potential impact, and whether the development efforts will truly benefit local residents.

Culpeper County, Virginia, is setting an example through strategic planning and partnership with DataBank, a data center provider seeking to expand responsibly in Virginia. The leaders of Culpeper County developed the Culpeper Tech Zone (CTZ), a model that addresses community concerns through intentional zoning, infrastructure planning, and workforce development.

DataBank’s approach to working with the CTZ exemplifies how the right partnership can deliver economic benefits that extend beyond tax revenue while keeping communities top of mind. This collaboration shows how municipalities and private partners can align on shared values to achieve responsible growth.

Strategic Planning in a Community

As data center development accelerated across Northern Virginia in the early 2020s, county leadership recognized an opportunity but needed to address the community’s concerns about scattered development and the overall impact of new infrastructure.

The strategic decision was to welcome data center development in a campus-style setting where infrastructure alignment made practical sense.

“We learned quickly that we didn’t want power lines crisscrossing throughout the entire county,” explains Bryan Rothamel, Economic Development Director for Culpeper County. “By identifying one electrical corridor and concentrating development where our community college and Route 29 already existed, we could centralize the infrastructure and maximize possible synergies.”

Culpeper identified a target area in close proximity to existing electrical service. It was also located along US Route 29, a four-lane highway for valuable transportation capacity. In addition to this, the adjacent community college and high school technical education program created natural workforce development opportunities.

The result was the CTZ, a 1,000-acre plot representing 0.4% of the county’s total land area. This strategic concentration meant the county could clearly communicate its vision: development in this specific area. By defining where development should happen, Culpeper could have productive conversations with potential partners rather than debating fundamental land use questions with every inquiry.

“Some companies wanted massive single-operator campuses,” says Rothamel. “That didn’t align with our vision for a concentrated technology zone serving multiple partners. We were comfortable saying no to proposals that didn’t fit, which actually strengthened our relationships with companies like DataBank that did.”

Why DataBank Chose Culpeper

From a private sector perspective, several factors made Culpeper attractive. DataBank already operates four data centers in Northern Virginia, making Culpeper’s location strategically valuable.

When data centers are located within close geographic proximity, they can efficiently sync data between facilities, provide backup during outages, and maintain redundant systems without the latency issues that come with longer distances. These cluster advantages make the infrastructure more reliable and responsive than isolated facilities.

Beyond geography, the CTZ offered something increasingly rare: process clarity. The streamlined approach from concept to groundbreaking eliminated uncertainty that typically plagues development timelines. When a developer asks, “Where would you like us to go?” and receives a clear answer, planning becomes dramatically more efficient.

The CTZ model demonstrated that local government had thought carefully about its infrastructure needs, community concerns, and long-term vision. This was proactive economic development, not reactive zoning. For companies evaluating multiple locations, this strategic thinking is the sign of a reliable partner, not just a jurisdiction that might shift direction mid-project.

Economic Impact: Beyond Tax Revenue

Data centers contribute substantial tax revenue during both construction and operational phases. However, focusing solely on tax dollars misses the larger, longer-term economic impact these facilities create.

The most direct benefit is employment. Data center operations need skilled technicians to maintain systems and keep facilities running 24/7. During the construction phase, projects employ local contractors, electricians, HVAC specialists, and construction workers—creating jobs long before the facilities open. These construction workers and ongoing employees also support local businesses, from coffee shops and restaurants to gas stations and retail stores, creating a ripple effect throughout the community’s economy.

Additionally, having the community college and educational institutions right in the CTZ creates immediate workforce development opportunities. The college can build training programs around what employers actually need, not abstract courses that don’t match real jobs. Students can meet potential employers while they’re still in school rather than hunting for jobs after graduation.

These connections are already happening. Companies operating in the CTZ are reaching out to meet with students and actively hiring from local educational programs. This creates a pipeline where young people can build careers without leaving the county, addressing the challenge many rural communities face when trained residents must relocate.

Community engagement extends beyond workforce development to include partnerships with local organizations, support for educational initiatives, and participation in county activities. Tax dollars generated in the CTZ can fund services, infrastructure, and programs throughout the county’s 383 square miles, allowing development and preservation to coexist.

A Framework for the Future

The Culpeper story extends beyond one county or one data center provider. It demonstrates what becomes possible when communities lead with clarity and companies respond with real partnership.

“We started every meeting with partners saying we’re looking for the best example of data center cluster development in the world,” Rothamel notes. “Working with DataBank, we’re building that together. It’s proof that rural communities can define their own terms for growth.”

The formula is straightforward: define your vision, plan deliberately, and choose partners accordingly. The Culpeper and DataBank relationship proves it’s more than possible – it’s achievable.

AI-driven power demand refers to increased energy requirements from AI workloads, which data centers manage through efficiency improvements and infrastructure planning.

A version of this article first appeared in The AI Journal. 

The current AI boom has quickly created unprecedented demand for data centers, and with it, exponentially higher demands for power.

The estimates and trends would have seemed impossible just a short time ago. Recent research projects that power requirements from AI data centers could grow more than 30 times by 2035, with some facilities scaling from traditional 10-50MW power requirements to 100-500MW or more.

Additionally, NVIDIA and other GPU chipmakers have released product roadmaps calling for 1MW per server rack, raising new questions related to power and cooling:

While many of today’s headlines seem to paint a gloomy picture of these new power demands —and data centers’ role— a closer look shows that there’s no reason to panic.

Innovative Power Strategies Taking Hold

The industry isn’t waiting for these problems to solve themselves. Data center operators are already rolling out proven technologies that tackle power constraints head-on. These new innovations are proving to be effective responses to growing power demands and are helping prevent what could become an infrastructure crisis.

A 2025 Duke University study found that if data centers can be flexible for just 0.25% of their operating time —representing just 22 hours per year— the U.S. grid can accommodate 76GW of new data center load without building new power plants.

Forward-thinking data center operators are now implementing a wide range of new strategies to deliver exactly this type of grid-connected flexibility.

Energy Storage Systems

For example, advanced storage technologies are quickly revolutionizing how data centers manage power fluctuations and grid interactions. Beyond traditional lithium-ion systems, data centers are implementing diverse storage solutions, including alternative chemistries like aluminum-ion and sodium-ion systems, solid-state technologies that offer enhanced safety profiles, and flow batteries that provide scalable, long-duration storage capabilities.

Additionally, emerging approaches such as thermal storage systems and gravity-based storage are also gaining traction for their ability to store large amounts of energy with minimal degradation over time.

These systems excel at managing the daily “duck curve” of renewable energy availability: storing excess solar generation during peak production hours and releasing it during evening demand spikes. This capability allows data centers to maintain consistent operations, reduce strain on the grid during peak demand periods, and maximize their use of renewable energy sources.

Virtual Power Plants

Additionally, virtual power plant (VPP) networks are transforming facilities from simple energy consumers into active grid participants. By connecting their storage or backup generation systems into coordinated networks, data centers provide grid stability services, participate in demand response programs, and optimize renewable energy utilization across entire regions.

DataBank has implemented VPP technology in California, demonstrating how storage systems can operate profitably while supporting renewable integration and reducing peak demand on fossil fuel plants. This example provides valuable insights into how data centers can contribute to grid stability while maintaining operational excellence.

While these flexibility innovations show significant promise, they also highlight the fundamental infrastructure challenges that must be addressed to fully unlock AI’s potential.

Why We Need to Focus on the Grid

While so much attention tends to focus on whether the U.S. can generate enough power for AI’s explosive growth, the real challenges lie in two areas: making data centers more flexible so they can work with existing grid capacity and upgrading transmission infrastructure to deliver power where it’s needed.

Flexibility Over Generation

The data center industry isn’t necessarily facing a fundamental power generation problem. The U.S. grid already maintains substantial reserves, with operators typically utilizing only about 50% of available generating capacity by design. This conservative approach ensures adequate power remains available to meet peak demands that occur during a small fraction of the time, such as the coldest winter mornings and the hottest summer nights.

This creates real potential for new operations strategies, such as demand response capabilities, energy storage systems that shift power usage to off-peak hours, and workload management that schedules non-urgent computing during times of grid abundance. Google’s recent agreements with Indiana Michigan Power and Tennessee Valley Authority represent the first implementations of data center demand response targeting machine learning workloads specifically.

In these cases, Google will temporarily reduce or pause energy-intensive AI and ML tasks during peak demand periods, enabling their data centers to operate as more flexible resources that don’t further strain the electrical grid.

The Transmission Bottleneck

While flexibility solutions help, the real constraint isn’t power generation capacity but utilities’ ability to deliver electricity where it’s needed. Transmission and distribution limitations, rather than generation shortfalls, are creating the bottlenecks that utilities must address through infrastructure investments.

The numbers tell the story. For example, Dominion Energy’s transmission constraints have created multiyear backlogs for Northern Virginia data centers, with relief coming only when new grid infrastructure comes online in 2026. This pattern repeats across major markets, where planning and construction of new transmission lines can take five to ten years, depending on regulatory approval, permitting, and right-of-way acquisitions.

In California, PG&E has stated that new substation work for large load interconnections may take five years or more. Additionally, all six major regional grid operators outside of Texas have reported they cannot meet the Federal Energy Regulatory Commission’s (FERC) deadline for critical transmission upgrades.

These timelines simply don’t line up. While data center developers can build facilities in two to three years, the time required to complete associated interconnection studies and infrastructure upgrades usually spans four to eight years. Goldman Sachs Research estimates that $720 billion of grid spending through 2030 may be needed, but the infrastructure development cycle simply can’t keep pace. This explains why operators are increasingly forced into expensive behind-the-meter (BTM) solutions rather than waiting for grid connections that may take half a decade or more.

Next Steps for Stakeholders

Moving forward means addressing both sides of the equation: transmission bottlenecks and data center innovation working together. While grid modernization, faster permitting, and better transmission technology are essential, data center operators are also stepping up with solutions that reduce grid dependence. The combination of smarter infrastructure and more flexible facilities will unlock sustainable AI growth.

The industry response shows that progress requires both utility infrastructure and data center adaptation. As the Duke University research demonstrates, even minimal data center flexibility can accommodate massive new capacity without additional generation. Meanwhile, behind-the-meter power solutions, clean energy adoption, and strategic facility placement are creating new pathways that work around current grid limitations.

Making this work requires coordination on multiple fronts. Utilities need to update their planning and speed up approvals while data center operators continue to utilize flexible power solutions such as energy storage, demand response, and behind-the-meter solutions. Regulators should prioritize transmission investment while encouraging data center innovations that support grid stability. Both sides are bringing solutions to the table.

The data center industry is adapting with proven technologies while pushing for better grid infrastructure. This isn’t a crisis spinning out of control but more of a two-sided infrastructure challenge where both transmission upgrades and smarter data center design will create more efficient and sustainable digital infrastructure. The question isn’t whether one side can solve this alone, but how quickly both transmission modernization and data center innovation can work together to power AI growth sustainably.

Corporate social responsibility initiatives reflect how organizations contribute to social causes and community support beyond core business operations.

Dozens of DataBank employees across eight data centers recently came together for a meaningful initiative: writing letters of appreciation and encouragement to U.S. military service members. Teams in Atlanta, Denver, Dallas-Fort Worth, Newark, Houston, New York, Orangeburg, and Salt Lake City participated in the effort, collectively sending more than 500 letters to troops serving at home and abroad.

The letter-writing campaign gave DataBank team members an opportunity to express gratitude and provide a personal connection to those serving in the military. From heartfelt thank-you notes to words of encouragement, each letter represented a small gesture with the potential to make a big impact on service members who are far from home.

DataBank's Letters for the Troops initiative

The result of DataBank’s company-wide initiative: more than 500 letters ready to be sent to troops serving at home and abroad.

Research shows that letters from home significantly boost morale and mental wellbeing for deployed troops, helping to reduce feelings of isolation during extended deployments. These personal connections remind service members that their sacrifices are recognized and appreciated by the communities they protect.

Each DataBank location approached the initiative in its own way, with some teams gathering during lunch breaks to write letters together while others set up stations where employees could contribute throughout the week. The collaborative effort not only supported military personnel but also strengthened bonds among DataBank team members who came together for a shared purpose.

Handcrafted with Care

While DataBank’s letter-writing campaign brought teams together across multiple locations, one employee has been quietly supporting the troops through his own unique contribution for years. Stephen Maiorca, Senior Security Architect at DataBank, has been handcrafting wooden pens for deployed service members for deployed service members since 2017. His efforts began when he first participated in a “Turn the Pen for the Troops” Veteran’s Day event as a local Woodcraft store with his teenage son.

Handcrafted wooden pens for the troops

Handcrafted wooden pens made by a DataBank employee for deployed troops.

The hobby began as a way to honor the military veterans in his own family—his father, who served in the Army; his grandfather, who was in the Navy during World War II; and his uncle, who served in the Air Force.

Over the years, he’s refined his craft and expanded his contributions. Each pen requires about two hours of work over a 24-hour period, including sawing, drilling, gluing, turning on a lathe, sanding, polishing, and final assembly. He donates these handcrafted pens to For The Troops, a local organization in Simi Valley, California, that creates care packages for deployed service members.

“I don’t always hear back from the troops, but when I do, they’re extremely grateful,” said Maiorca. “I actually ran into one of the soldiers I’d sent a pen to, and he told me it helped him keep in touch with his wife and kids. That’s why I do this—to help the people who are protecting our freedoms and way of life.”

Supporting Those Who Serve

The letter-writing initiative reflects DataBank’s commitment to supporting the communities where its teams live and work. Whether through organized company efforts or individual acts of service, DataBank employees continue to find meaningful ways to give back and express gratitude to those who serve.

For service members who are far from home, these letters and handcrafted gifts serve as reminders that their sacrifices are seen, valued, and deeply appreciated.