Systems architect, founder, and full-stack engineer building AI-native platforms at the intersection of automotive, computer vision, and systems software. Based in Houston, Texas.

Senior Software Engineer @ NexIT Group (June 2026 – Present)
Designing and shipping client applications, websites, and internal tooling — including a cross-platform Rust vulnerability detection & patch-management agent, React + Three.js experiences, and AV / UI systems work.

Owner / Operator of @RapidFire-AutoCare and continuing the deep-tech exploration that was expanded from NextGen AutoGlass to include Ubiquity's Automotive Datasets and now that development continues through @Ubiquium AI.

Right now I'm focused on:

• Kernel-to-cloud systems, AI agents, and production CV pipelines
• Open-source developer tools (NovaDiff) and vehicle diagnostics (rlink-j2534-macos)
• Spatial / 3D web components (anomalous-three.js-components)

I grew up around aircraft, sensors, and machines that fail loudly when you get the details wrong. That obsession turned into software early: by high school I was shipping real client work and attempting something most people would call impossible.

Early paid / contract work

• Texas Direct Auto (2013) — Built the inventory instant-search experience (AJAX / jQuery / T-SQL) that stayed in production until the post-acquisition migration to Vroom.
• ION Geophysical (2014) — Software engineering intern on Prometheus / BASIN SPAN: Doxygen docs across 1M+ lines of C++, CUDA exposure, Perl automation, and a KDE Marble–based trade-show geospatial viewer.
• NexIT Group (2015 – 2017) — First full stretch as a professional engineer: websites, applications, AV deployments, access control systems, networking & cyber security, and the design & development of systems for clients.

xyOS / Xymbi (2012 – 2016) — During high school, I founded Xymbi and engineered xyOS: an ambitious cross-platform Linux-based operating system built to unify application execution across Windows, macOS, Android, and iOS—long before cross-compatibility solutions like WSL, Proton, and Rosetta 2 became household names.

Technical Architecture:
xyOS combined and extended open-source compatibility layers—leveraging Wine for native Win32/Win64 application execution, Darling for Mach-O/macOS compatibility, and in-progress ART/Mach support for running Android and iOS apps. With custom kernel modules, I introduced dynamic system call translation and binary interface mapping, allowing simultaneous execution of Windows PE, macOS Mach-O, Android ART/Dalvik, and iOS ARM64 binaries. At the kernel level, xyOS implemented syscall proxies so that Windows DLLs, Cocoa frameworks, Linux shared objects, and even multi-architecture mobile binaries could coexist on the same runtime. File system interoperability was achieved by bridging NTFS, HFS+, ext4, and early APFS support through a custom FUSE-based VFS, enabling seamless access for applications expecting vastly different host environments.

The project pushed beyond desktop OSes:

• I integrated and patched major game console emulators (PCSX2, Dolphin, XQEMU/Xenia) for native support, using API abstraction layers to dynamically translate DirectX, Metal, and OpenGL/Vulkan calls.
• A registry abstraction emulated Windows NT-compatible keys alongside macOS .plist config morphing, supporting multi-environment application state management.
• Dynamic dependency and driver resolution permitted DirectX and Metal translation via a unified OpenGL/Vulkan driver model, including QEMU-based syscall translation for non-native code paths.
• All of this ran directly on hardware—decades ahead of container-based or virtualized approaches.

Relevance and Significance:
xyOS was pioneering in its vision and execution, enabling cross-platform application and game execution without containers, VMs, or conventional virtualization. It anticipated the rise and value of cross-compatibility frameworks like Proton, Rosetta 2, WSL, and the broad shift to containerization as the industry norm. xyOS aimed to break down the lock-in and friction inherent to proprietary OS silos, giving developers and users a unified, consistent experience—years before the industry standardized on easy app portability.

Why it Ended:
The project reached a painful inflection point in 2016. While pitching xyOS to investors at NextGen Summit (Austin), a catastrophic remote breach led to the total wipe of our backend servers and my primary development systems—terminating active development and erasing most of the original source. Simultaneously, the OS landscape changed fundamentally:

• Microsoft introduced AppX/MSIX and virtualization for Windows, with containerization and WSL paving the way for easier, supported cross-platform execution within Windows itself.
• Apple locked down iOS and macOS, introducing rootless and SIP protections, restricting system-level access and effectively ending the viability of third-party system-call translation on Apple platforms.
• Mainstream containerization emerged (Docker, LXC), sidestepping the original kernel-level compatibility challenge by enabling secure, reproducible lightweight VMs, further eroding the need for highly invasive kernel modifications.
  As the major vendors closed their ecosystems, the overhead and risk grew exponentially for independent OS projects like xyOS.

Takeaway:
Though the project was ultimately shelved, the technical and architectural lessons from xyOS—kernel engineering, distributed system design, deep compatibility translation, and security—deeply shaped my understanding of resilient platforms and the realities of building under fast-moving industry constraints. xyOS set the stage for everything I’ve built since.

Before I automated glass and diagnostics, I did the work by hand: Discount Tire, Safelite (mobile → lead tech → market ops), and multiple body shops & automotive service centers assuming both the roles of technician at various levels and Service Director and Operations Coordinator respectively at my highest positions in the automotive corporate sphere. That apprenticeship is why my software maps to real shop constraints — parts, SLAs, ADAS, techs in the field, and customers who need their cars back today.

Studied Aeronautical Science at Embry-Riddle Aeronautical University (2016 – 2018). -> Dropped Out

Founded NextGen AutoGlass and rapidly scaled an AI-powered automotive glass and procurement platform—hitting $2.7M ARR in its first year and servicing 10,000+ vehicles throughout Texas metro areas (DFW, Houston, San Antonio, Austin). NextGen offered 24/7/365 fleet support for enterprise clients such as Hertz, all while maintaining a 0% warranty claim rate under NHTSA / FMVSS standards—an achievement in a field often plagued by rework and quality failures. I served as CEO under existing interim CEO Paul Stone (a conservative American) as we built the operation out of Downtown Houston HQ at 405 Main Street, Suite 300D.

In July 2020, NextGen entered a high-stakes partnership with Hertz at Houston Hobby Airport to clear a pandemic-fueled backlog of vehicles needing glass repair, then expanded to George Bush Intercontinental (IAH) and DFW International. At that stage, NextGen was servicing Hertz with zero operational issues—units out, quality held, demand real.

The political context mattered. In mid-2019, while we were still building toward that Hertz relationship, Goldman Sachs CFO Stephen Scherr donated to Joe Biden’s presidential campaign—later clarified publicly with a partial refund on the contribution. Parallel tracks: Wall Street capital aligning with a presidential bid, and a Texas fleet vendor quietly keeping rental cars glass-ready under a NET 30 contract.

When Biden was elected, NextGen’s downtown Houston office streamed live general-election vote data to our machines through the night—tapping the same NYT-referenced Dominion / Edison voting API feeds other independent developers were watching in real time. We recorded what we saw on video and saved raw JSON copies instead of trusting what cable news displayed when “pipes burst” and “stop the count” events hit PA and GA mid-night, while our systems kept writing history to disk. That archive and analysis later became public as Matrix-Red-Pill-2024—an attempt to preserve primary-source timelines when institutional narratives and live data diverged. Houston was boarding up; employees came to the office armed over fear of the political climate; COVID was still raging. We kept the feeds running anyway. Now let me ask this, if they jailed non-violent citizens who toured the capital on January 6th. What do you think they did with the location, business and personal information behind the IP addresses that conducted GET REQUESTS against the voting data without being on a media IP White List?

Technologically, NextGen pioneered several significant advances:

• FastLane — The first-of-its-kind, a patent pending drive-through camera system for automated damage assessment, part selection, diagnostics, recalibration, and instant service procurement.
• NextGen FrontLine — The world’s first on-demand automotive tech app: empowering technicians with full lifecycle control (self-dispatch, navigation, onboarding, documentation, flexible scheduling, instant payouts).
• NextGen Realm — A unified, serverless cloud platform (Bref PHP on AWS Lambda), acting as the nucleus for all operations, decision logic, and field-to-cloud automation—governed by Iris (Intelligent Real-Time Invocation System), an AI entity interfacing at the kernel level to drive order handling, vision tasks, logistics, and automated workflows via TesseractOS Core Runtime.

Those systems were built for pandemic reality: remote work, automated diagnostics, live inventory, and hands-off scheduling. My team and I made 24/7 fleet glass possible when legacy competitors still ran on clipboards and phone trees. Stonewalled by Corporate / Political Bias & Influence.

Then the Hertz leadership stack shifted. Stephen Scherr was announced as CEO of Hertz Global Holdings. Strategy pivoted. Paul Stone reached out to try and warn me.

Black Tahoes rolled up with a Hertz VP who had zero interest in meeting or discussing anything with me—even though I was on site. They manufactured an issue with NextGen working the parking garage at IAH and escalated it through Lucian Manning (IAH Maintenance Management). Simultaneously, Hertz began actively soliciting my employees while we continued to crank out units and demand surged as the semi-post-COVID market settled back toward pre-COVID norms.

Then the money stopped behaving like a Fortune 500 NET 30 relationship:

• Hertz AMEX payments bounced non-stop. Expansion was promised. Settlement of balances was promised.
• I was forced into merchant cash advances (MCAs) out of NYC to keep payroll and parts moving—capital that assumed revenue would hit the account on schedule.
• Hertz drifted to roughly 290 days past due on AR / open invoices against a NET 30 contract. MCAs came to collect on money that never arrived.
• Lawsuits started flying at NextGen. In a meeting with Lucian, he told me the decision came from the very top: NextGen AutoGlass had been put on an “Executive Payout Hold.” He told me privately that decision likely came from the very top of the Organization. Scherr himself.

NextGen lost the contract after we halted service for non-payment. Multiple six-figure emergency capital injections, ex-employee shell entities skirting controls, chargebacks, Hertz’s own bankruptcy-era chaos, and a ~$215K unpaid-invoice backlog finished what the payout hold started. By Christmas 2021 the operation was on emergency measures.

I wrote about what happened. I made a LinkedIn post tagging a publicly traded Fortune 500 company (Hertz) and got 14 views, half of which were Hertz Executives and US Government related. I realized I'm not visible, not amplified, "the algorithm" filters me out. I was declared a ghost. That was the start of the shadow-ban phase.

Then the part that wasn’t corporate: the Pfizer COVID mRNA vaccine that was sold as “safe and effective” and treated as mandatory for people in my position induced myocarditis, a recurrent SVT loop, a heart attack (left-ventricle scarring), and subsequent cardioversion, leaving mitral, tricuspid, and pulmonic valve regurgitation. The ghost went on bed rest 2022–2023.

Despite its operational failure, NextGen left a mark:

• Technological Leadership: FastLane and FrontLine introduced industry-firsts for AI-enabled auto glass—automation, mobile workforce tools, real-time claim / field processing.
• Market Impact: At peak, NextGen kept essential fleet vehicles glass-ready through the pandemic when large parts of auto services were paralyzed. The serverless, AI-driven backend set a technical bar competitors had to chase.
• Political / Institutional Lesson: The unwind was not only “COVID cashflow risk.” It sat at the intersection of enterprise power (Hertz under new political-adjacent leadership), payment as a weapon (“Executive Payout Hold”), information control (election-night raw feeds vs. media narrative; LinkedIn erasure), and the personal cost of mandatory public-health policy on a founder already under siege. The Matrix-Red-Pill-2024 archive exists because we refused to outsource memory of that night to whoever owned the chyron.
• Cross-industry Legacy: Hard lessons in enterprise deal structure, AR discipline, and workforce enablement fed directly into Ubiquity’s multi-domain, AI-powered platforms—and into a permanent refusal to build systems that depend on a single counterparty’s goodwill.

Peak valuation post-MVP and pilot ops: >$25M (March 2024). Forbes Business Council member (2020–2022).

NextGen did not die of a bad product. It died of unpaid invoices, institutional pressure, and a body that finally gave out. The tech—and the cautionary record of how power actually moves—survived into what came next.

As Co-Founder & Principal Systems Architect, I rebuilt the stack into a connected vehicle service ecosystem:

Patent filed: Systems and methods for automating vehicle identification, damage assessment, and service procurement using LiDAR, AI, and camera array technologies (No. 63/699,790).

I started working directly with owner Hassan, who saw—through personal exposure and live demonstration—the potential of the agentic AI systems I’d built in auto glass. He wanted that same functionality applied to his vehicle service and auction-vehicle restoration businesses.

As Service Director, I wore both hats at once: building and deploying the stack, and running the floor. On the systems side I shipped an Azure + Docker enterprise CMS for diagnostics / repair tracking, AI scrapers for wholesale parts (AutoZone, O’Reilly, XLParts, …) over WebSocket, the PitCrew iOS field app, and a GoDaddy → Microsoft 365 + SAML SSO migration—while expanding software dataset ingestion capabilities beyond what I’d had in the NextGen era. On the shop side I was servicing vehicles, overseeing all on-site technicians, triaging jobs, diagnosing vehicles, ordering parts, calling customers, getting approvals, and walking findings / estimates.

He underpaid me. Partnership promises never materialized. The relationship stayed one-sided: I built, deployed, and carried the operational load; he did not pay or hold up his end of the bargain. I took my intellectual property and left.

Returned to NexIT Group as Senior Software Engineer while standing up the next chapter under RapidFire / Ubiquium: automotive care, systems software, and open tools that anyone can run.

Languages — C#, C++, Rust, Python, Ruby, Node.js, PHP, Java, Kotlin, Swift, Objective-C, Perl, HTML/CSS, GLSL, TypeScript, Go, Shell
AI & Vision — Agents / LAMs, LLM fine-tuning (LoRA / QLoRA), Hugging Face, PyTorch, OpenCV, ALPR, 2D/3D segmentation, object detection, LiDAR, spatial computing, USDZ
Cloud & Systems — AWS, Azure, GCP, OpenStack, Docker, Kubernetes, CI/CD, serverless, Linux kernel, CUDA, Windows / macOS / iOS / Android
Data & Realtime — MS SQL, MySQL, MongoDB, DynamoDB, CockroachDB, Aurora, REST, Webhooks, WebSockets
Security & Ops — Tenable / Nessus, Vicarius TOPIA, Nmap, SAML SSO, MFA
Design & Spatial — Adobe CC, Figma, Unreal, Cinema 4D, After Effects, Three.js, Lottie

• LinkedIn: linkedin.com/in/anomaloushuman
• GitHub: github.com/anomaloushuman
• Web: rapidfire.care

Always building. If you’re working on systems software, automotive AI, or tools that respect the metal — let’s talk.