Throughout my career, I’ve independently managed all technical requirements for my projects, each of which has successfully launched to market and performed well in production. My self-taught background allows me to build entirely custom solutions, delivering precisely what business stakeholders need. I’m highly driven, collaborative, and passionate about my work.
Having run my own businesses, I bring a strong understanding of broader objectives, enabling me to quickly develop products that align with a company’s larger vision. While I set high expectations for myself and my colleagues—sometimes challenging to others—I actively help team members succeed in their roles and have guided many to improve over time.
My role as Founder of Slicie encompasses every aspect of the business, as I serve as the Manager of this LLC. I handle all operational and technical responsibilities with the financial backing of my co-founder, Rick Garcia.
Slicie was created as an alternative to competing in the hyper-competitive cPanel hosting market that Crucial occupied. Crucial’s business declined because it focused on Magento hosting, which was ultimately overtaken by simpler solutions such as Shopify and WooCommerce, in addition to the burdensome transition to Magento 2.
Slicie is fundamentally a platform that provides Automatic Scaling hosting based on a utility billing model. Nearly everything in Slicie was developed from scratch, aside from essential services like libvirt/KVM, Caddy, NGINX, RabbitMQ, Galera Cluster, and similar components.
Achieving on-demand scalability for virtual machines required building a custom storage platform that I believe is unique in the industry. Customers can pay for as little as 64 KiB of space while being allocated more than 16 TB (up to 256 TB). From a performance standpoint, local NVMe storage is used for hot (frequently accessed) data, while remote lower-endurance NVMe connected via NVMF handles less frequently accessed data. Rather than a simple caching layer, data is moved in real time between tiers—promoting and demoting data between these layers—using a cascaded storage model. This, along with real-time billing, required developing an extensive Linux kernel module in C, interacting with a high-availability user-space cluster that allocates 128 MiB “regions” of data per customer.
The storage platform supports real-time statistics reporting (measured in microseconds) for each layer, live migration of data blocks via NVMF, real-time image import and export, instant creation and restoration of backups, real-time promotion and demotion of data between performance tiers, and a pay-for-what-you-use model with discard support (allowing customers to use fstrim to lower usage). It also offers defragmentation, region reclamation, and live migration between data centers (by tagging 64 KiB blocks with a data center ID). This custom storage system has operated in production without failure for several years, having been extensively tested with over a petabyte of I/O.
On Slicie, virtual machines scale up or down automatically. They are continuously monitored in real time, with memory allocated based on utilization—including caching considerations—and CPU cores hot-plugged as needed. Slicie also features instant backups that leverage deep integration with the storage drivers, enabling backups to be mounted on existing servers or cloned to run alongside the original VM.
There are many more features, but for brevity, you can explore slicie.com to see how the platform continues to grow. Beyond the technical work, I also manage the overall business and marketing.
My first project at Crucial Web Hosting was creating a real-time statistics program to monitor all of their servers (around 300), intended for both internal use and customer-facing access. I completed this in the first month, and it remains in use to this day.
In 2011, Crucial was paying IBM approximately $36k per month for dedicated cPanel hosting servers. They were struggling with significant performance issues on bare metal, which led to both constant performance bottlenecks and severely underutilized servers (the worst of both worlds). Crucial’s support for Magento hosting on Virtuozzo was noteworthy, especially since Virtuozzo at the time did not provide effective isolation or control over multi-tenancy.
I moved Crucial away from SoftLayer and Virtuozzo by migrating them into a cage I set up at PhoenixNAP (located near where employees lived). Due to Crucial’s size (about $1M per year in revenue), we needed five cabinets of physical servers. To reduce costs, I built all of these servers by hand using commodity hardware (primarily E3-1270s, with a few E5s). I learned the necessary networking skills to implement a leaf-spine topology with bonded links on the Xen hypervisors for active-passive failover. We used Cisco switches with HSRP for gateway redundancy and BGP connections to upstream peers, along with public IP space acquired from ARIN.
On the billing side, I migrated our customer-facing integration away from ModernBill, though we continued to use ModernBill internally because the team was accustomed to it. For security, I replaced the underlying billing mechanisms—tokenizing cards with Authorize.net—and revamped the way ModernBill communicated with our infrastructure.
Although we continued using Kayako for the helpdesk, I customized the on-premises version to integrate in real time, pulling customer details about billing and servers for easier access by our support team.
I also developed a custom backup system that performs daily backups on all servers, retaining historical backups for up to 30 days. We turned this feature into a “premium backups” product, allowing customers to manage their backups directly through cPanel. This offering quickly became popular among our clients.
Over time, I’ve taken on all aspects of running the business. I maintain close relationships with our customers through support, represent the company in legal proceedings, handle hiring and terminations, and generally manage every other facet of operating a small business.
At HostGator.com, I was in charge of overseeing all aspects of software development for the company.
I developed a custom helpdesk to replace Kayako, which at the time was used by roughly 500 employees. I completed this replacement in about one month. Aside from the requirement to provide instant page loads on both the customer and staff sides, the system included a real-time locking mechanism to prevent multiple staff members from accessing the same ticket (unique in 2008), a method to automatically assign staff to the next ticket based on prior replies (matching them with the “best” next ticket), and a “reward” system for support staff to earn bonuses based on game-like metrics—such as streaks of quality responses and five-star reviews—along with a leaderboard that displayed employee performance by department and team.
I replaced the existing billing system (ModernBill) in about one year, primarily due to persistent problems with ModernBill. At peak hours, ModernBill’s page load times were 40–60 seconds, whereas my solution (GatorBill) reduced them to under one second on a single four-core server. Some features of GatorBill included integration with my other custom software, more intelligent provisioning of up to a thousand new customers per day, integrations with VPS, uniquely intelligent fraud prevention, and deep integration with my custom affiliate system. Replacing ModernBill was particularly challenging because it was closed-source and had an unwieldy database schema, but after transitioning to GatorBill, we actually increased revenue due to purchases that had been stuck in ModernBill.
I also replaced HostGator’s chat system (LivePerson) by distributing hundreds of live chats to employees based on supervisor-configured settings (such as how many chats each employee could handle). This was uniquely challenging at the time because AJAX support was in its infancy, and LivePerson relied on refreshing iFrames, which we had to fall back on for roughly 10% of browsers (particularly early Internet Explorer). I collaborated with a team member I managed, who created a C++ interface for the employee side to offer a desktop application (this was in 2008, before browser implementations were robust enough).
I created a custom affiliate system that processed up to $500k per month in payments over a weekend and significantly reduced fraud (by around 20–30%) through an integration with GatorBill. I also developed internal software to monitor approximately 6,000 physical servers at SoftLayer (now IBM Cloud), as well as a template system and an affiliate system that required a custom multi-threaded HTTP server capable of serving billions of requests efficiently on commodity hardware.
Beyond creating software, I managed a small team of developers responsible for replacing the company’s VoIP phone system and building a custom cPanel theme (since HostGator used cPanel at the time). I was additionally responsible for representing HostGator during discussions to sell the company, though I left the company due to a falling-out prior to its eventual sale to EIG.