Systems software engineer & security researcher focusing on low-level Windows internals, native runtime architectures, and AI-augmented engineering.
This space serves as a central index for my engineering archives, technical research, and system utilities.
π‘ Repository Notice & Licensing
Core engines and proprietary research projects are maintained in private workspaces. Architectural overviews and technical articles are documented via my blog and documentation site. Upstream forks adhere to their original open-source licenses.
- Low-Level & Systems Programming: Assembly (MASM, NASM, FASM), C/C++, Object Pascal (Delphi, FPC), Win32 Internals, Memory Management, PE/Binary Analysis.
- Security & Runtime Virtualization: Custom Virtual Machine (V-CPU) design, anti-tampering/anti-debugging mechanisms, binary protection.
- High-Performance Pipelines: Zero-copy architectures, hardware-accelerated media encoding (NVENC / Media Foundation), network packet fragmentation.
- AI-Augmented Engineering: Agentic workflows, frontier LLM orchestration (Claude, Gemini), local model environments (Ollama), and custom developer tooling.
- Domain: User-mode code virtualization, binary obfuscation, and runtime integrity protection.
- Key Architectures:
- Designed a proprietary Virtual Instruction Set (VIS) and high-speed V-CPU interpreter.
- Implemented zero-copy virtual execution loader and selective dynamic virtualization.
- Built-in multi-layered anti-analysis mechanisms (hardware breakpoint detection, anti-debugging, PE header restructuring).
- Domain: Real-time screen capture and hardware-accelerated encoding pipeline.
- Key Architectures:
- Zero-copy desktop duplication pipeline combining Windows Media Foundation (MF) and NVENC.
- Optimized memory layout and double-buffering to minimize frame drop and capture latency.
- Domain: Low-level packet inspection and transport security.
- Key Architectures:
- Win32 network stack integration with DNS-over-HTTPS (DoH) resolution.
- Client-side packet fragmentation and socket-level filtering pipeline.
- Domain: Local LLM orchestration and structured consistency tracking.
- Key Architectures:
- Designed deterministic state-tracking pipelines integrated with local models and SQLite.
- Automated verification of plot consistency, context integrity, and cross-reference validation.
- Languages: C/C++, Assembly (x86/x64), Object Pascal, Python, TypeScript, Rust, Go
- Environment & Native Toolchains: Visual Studio, Windows SDK/WDK, GCC, Android Studio
- System Scripting & Automation: PowerShell, Windows Batch, AutoHotkey, AutoIt
- AI-Driven Engineering: Antigravity IDE, Cursor, Claude Code, Gemini CLI, Ollama, LM Studio
- Personal Site: gurupia.github.io
- Technical Writings: gurupia.dev
