Research and engineering software, developed in parallel with graduate research activity. All source code is available on GitHub.
Local LLM runtime written in Rust. Loads GGUF-format models, bridges to C++ inference libraries via FFI, and exposes an HTTP inference interface. Designed for cross-platform (macOS / Linux) reproducible local LLM execution with a focus on resource-constrained deployability.
Qt/C++ CLI-based OS internals learning project. Implements and visualises core OS concepts: process scheduling, memory management, and file system structures. Designed for systematic study of kernel architecture through hands-on experimentation.
C++ inference environment built on LibTorch. Implements a model loading pipeline, a lightweight REST API for inference serving, and experimental integration with GGUF/llama.cpp backends. Developed as a practical substrate for edge AI research.