Skip to content
View Rolaand-Jayz's full-sized avatar

Block or report Rolaand-Jayz

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Rolaand-Jayz/README.md

Rolaand Jayz

Jeffrey Orr Jr. | Software QA, Technical Investigation & Linux/Video Engineering

I investigate what software actually does, not just what it is supposed to do. My background combines enterprise IT support with independent software engineering and rigorous test work: reproducing failures, reading code and logs, isolating variables, challenging apparently successful results, and documenting findings so someone else can verify them.

My main areas are manual and exploratory testing, evidence-based regression checks, C++/Python investigation, Linux, GPU/video pipelines, and AI-assisted engineering with independent validation. I use AI tools to expand the scope of what I can work on, while retaining responsibility for requirements, technical decisions, testing, and acceptance.

QA case studies and sample test design → — public evidence of defect investigation, failed hypotheses, test scope, and repeatable checks.

Selected public work

Temporal Forge Player — closed FSR-era video-reconstruction research

A Linux C++/Vulkan/FFmpeg research project investigating how AMD FSR 4.1-style temporal reconstruction behaves on ordinary decoded video. The most relevant work is the test methodology: controlled experiments, source-cadence requirements, baseline comparisons, multi-frame evaluation, negative findings, and review of automated quality gates.

  • Reopened a previously passing quality gate after visible lattice corruption appeared; retained and reclassified invalidated results.
  • Used evidence from repeated experiments to close the FSR-centered approach in September 2026 rather than claim a quality advantage that was not established.
  • Read the closure and evidence ledger.

AMD Video Enhancer (AMD-VE) — Linux-first AMD video application

Experimental C++ application integrating FFmpeg, ROCm/HIP, MiGraphX, and Vulkan. Testing went beyond the application's UI into runtime dependencies, model execution, GPU behavior, benchmarking, fallback paths, and reproducible release documentation.

The investigations prompted four submitted upstream MiGraphX optimization pull requests: #4707, #4708, #4709, and #4710. Benchmarks are workload-specific; this is not a claim of universal speedup or upstream acceptance.

FSR 4.1 Static Reverse Engineering — reproducible research

Analysis of shader passes, binary resources, dispatch structures, neural-weight data, and resource bindings using Python, C, disassembly, and comparison tools. The repository deliberately distinguishes verified static facts, inferred behavior, and runtime behavior not yet observed.

Current private work

ANVIL / FORGE: ongoing from-scratch video-reconstruction R&D informed by the preceding research. The active repository is private; the public Temporal Forge repository remains available as an auditable record of the earlier experiment and the decision to pivot.

How I approach quality

  • Start from observable behavior: expected result, actual result, environment, reproduction steps, and evidence.
  • Test the affected scope: inspect changes and dependencies; use normal, edge, negative, fallback, and regression scenarios as appropriate.
  • Isolate causes: controlled comparisons, instrumentation, system logs, benchmarks, and cross-platform checks.
  • Verify independently: visual inspection can overturn an automated PASS; compiling is not the same as fixing.
  • Document uncertainty: retain failed tests, caveats, test conditions, and enough context for others to reproduce the result.

Professional background & tools

  • Enterprise support: Service Desk Analyst, Alpha Net Consulting LLC, through Capgemini for Progressive Insurance (April 2025–July 2026); Microsoft 365, application/hardware/access issues, BMC Remedy, incident resolution and escalation.
  • Customer operations: Former ResultsCX team supervisor; led and coached teams of approximately 15–20+ agents.
  • Technical toolkit: C++, Python scripting, JavaScript/Node familiarity, Bash, Git/GitHub, VS Code, Arch/CachyOS, Ubuntu Server, Windows, FFmpeg, Vulkan, ROCm/HIP, MiGraphX, Oracle Cloud, networking, and cross-browser diagnostics.
  • Certifications: CompTIA A+, Azure Fundamentals (AZ-900), Azure AI Fundamentals (AI-900), Linux Essentials.

Interested in manual/software QA, technical validation, video/application reliability, systems integration, and engineering roles where troubleshooting and empirical reasoning matter.

LinkedIn — Jeffrey Orr Jr.

Pinned Loading

  1. Temporal-Forge-Player Temporal-Forge-Player Public

    Experimental Linux/Vulkan video player adapting AMD FSR 4.1-style temporal reconstruction to ordinary decoded video — no frame interpolation.

    C++ 4

  2. RE-of-FSR-4.1.0-Upscaling RE-of-FSR-4.1.0-Upscaling Public

    Evidence-driven reverse engineering of AMD FSR 4.1: DXIL structure, neural weights, dispatch behavior, reproducible tooling and strict claim verification.

    Python 5 1

  3. AMD-VE AMD-VE Public

    AMD-first Linux AI video enhancer using C++, MiGraphX, ROCm, Vulkan and FFmpeg — with measured GPU optimizations submitted upstream to ROCm.

    C++ 10 1