Diyorjon Holkuziev: "Managing multiple cloud manually is impossible"

DevOps engineer who migrated legacy systems to AWS and Azure explains how his automation framework achieved 99.99% uptime and cut costs by over $200,000 annually

According to Gartner, 93% of enterprises now run multi-cloud strategies, yet many struggle with the complexity. With businesses increasingly dependent on cloud infrastructure for critical operations, system reliability has become a make-or-break factor for competitive advantage. The challenge isn't just adopting multiple cloud providers – it's making them work together seamlessly while cutting costs and improving performance.

Diyorjon Holkuziev has cracked this code at Brilom Inc., a technology services company specializing in cloud infrastructure solutions, where he built infrastructure automation that runs across AWS and Azure with 99.99% uptime. His migration of legacy systems saved the company over $200,000 per year while reducing deployment time by 70%. Working across Massachusetts, Illinois, and Florida, his solutions support healthcare and finance clients where downtime isn't just expensive – it's potentially catastrophic. What makes his approach different is the integration of DevSecOps tools like Vault and Snyk into CI/CD pipelines, cutting security audit time by 60%.

Your infrastructure runs across three states and handles healthcare clients where compliance is non-negotiable. What was broken with the old system that pushed you toward this multi-cloud approach?

We had too many single points of failure, and our healthcare clients couldn't afford the downtime we were experiencing. The old setup was essentially putting all our eggs in one basket with a single cloud provider. When that provider had issues, we had issues. Our clients in Boston and Chicago were getting frustrated because system outages meant they couldn't access patient data or process transactions. The breaking point came when we had a four-hour outage that cost one client thousands of dollars in delayed operations. I realized we needed redundancy not just within one cloud, but across completely different platforms. The solution I built treats AWS and Azure as one unified system while keeping the best features of each – AWS for our compute-heavy analytics and Azure for our Windows-based applications that integrate with Microsoft ecosystems.

Your infrastructure optimization resulted in over $200,000 in annual savings for Brilom. That's a significant number for infrastructure optimization. How did you achieve those savings?

The savings came from three main areas that most companies miss. First, I built algorithms that automatically move workloads to whichever platform offers better pricing at any given time. Our data analytics jobs run on AWS during their off-peak hours when compute is cheaper, while our customer-facing applications stay on Azure for better integration with our clients' existing Microsoft infrastructure. Second, instead of scaling entire application stacks like most companies do, I designed granular auto-scaling that only adds resources for the specific microservices that need them. Third, I implemented lifecycle management that automatically deletes unused resources and moves cold storage to cheaper tiers. Before this system, we were overpaying for resources we weren't using and running workloads on expensive infrastructure when cheaper alternatives were available. The $200,000 in savings represents about 35% of our previous cloud spending, and that's while actually improving performance.

Your background includes working at Nestlé in Uzbekistan and then the Greek International Food Market in Boston. How did that journey influence your approach to building resilient systems?

That transition taught me that systems need to work under completely different conditions than what you originally designed them for. At Nestlé, I was a production operator dealing with physical systems that had to run consistently in challenging environments. When I moved to Boston and worked at the food market while studying computer science, I learned how important it is for operations to continue smoothly even when everything around you changes – different suppliers, different regulations, different customer expectations. I apply that same thinking to cloud infrastructure. You can't just build for perfect conditions; you have to assume things will go wrong and build systems that adapt. My multi-cloud setup isn't just about redundancy – it's about having infrastructure that can handle regulatory changes, provider outages, or sudden spikes in demand without breaking. 

You integrated DevSecOps tools and cut security audit time by 60%. What security headaches were you dealing with before this integration?

Security audits were a nightmare because we had to manually verify compliance across different systems and platforms. Every quarter, we'd spend weeks gathering logs, checking configurations, and documenting security measures for our healthcare and finance clients. The problem with multi-cloud is that AWS security works differently from Azure security, so we were essentially managing two separate security ecosystems. I integrated Vault, HashiCorp's secrets management platform, for centralized secrets management and Snyk, a developer security platform, for automated vulnerability scanning that works identically across both platforms. Now, instead of manual checks, we have automated compliance validation that runs continuously and generates audit reports automatically. The 60% time reduction means what used to take three weeks now takes about one week, and we catch security issues in real-time instead of during quarterly audits. For our healthcare clients especially, this matters because HIPAA violations can result in massive fines..

You've worked across different countries and regulatory environments. How does your international experience give you an edge in helping American companies navigate global cloud deployments?

Having worked in Uzbekistan's regulatory environment at Nestlé and then adapting to American business practices gives me a unique perspective on how cloud infrastructure needs to function across different legal and cultural frameworks. When our American clients expand internationally or work with overseas partners, I understand the compliance challenges they'll face because I've lived through similar transitions myself. For example, data residency requirements vary dramatically between countries, and what works for HIPAA compliance in the US might not meet European GDPR standards.