I'm a systems-focused software engineer building compilers, low-latency engines, and performance-critical infrastructure in Rust.
- 🎓 Computer Science student graduated in July 2025
- 💻 Looking for hybrid Rust/C++ systems roles in London
- 📌 Strong interest in: infra, protocol design, AI tooling, trading engines, multi-agent systems
- 🔍 Long-term goal: Become a high-performance specialist in compilers, GPU runtimes, and low-latency systems
- 🧵 High-performance compilers and kernel fusion (e.g. ONNX/TorchScript)
- ⚡ Lock-free engines and real-time matching systems
- 🧬 AI infra, graph optimizers, and runtime tuning
- 🌐 Network protocols, concurrency models, and async runtimes
- 🤖 Multi-agent AI systems and orchestration frameworks
- 🖼️ Local-first applications with embedded ML and privacy-preserving design
- 💹 Trading infrastructure and financial systems engineering
| 🔧 Project | 🧠 Tech | 📄 Summary |
|---|---|---|
| Tectra | C++20, FlatBuffers, Lock-Free, Shared Memory |
High-performance trading infrastructure with market data feed handler, pre-trade risk engine, kill-switch, and deterministic replay; targeting >1M msgs/sec with microsecond-level latency |
| Image Browser | Rust, Tauri v2, React 19, SQLite, ONNX, CLIP |
Local-first Pinterest-style image manager with semantic search via CLIP embeddings, visual similarity recommendations, and fully offline architecture with complete user privacy |
| Xyntra | Rust, egg, wgpu, ONNX |
Kernel-fusion compiler pass for ONNX/TorchScript graphs, emits fused GPU kernels using WGSL/PTX |
| Zyphos | Rust, TCP, Threads, Concurrency |
Network protocol laboratory implementing HTTP servers from scratch in Rust; learning sockets, concurrency, parsing, and performance engineering through 30 structured milestones from raw TCP to HTTP/2, WebSockets, and QUIC |
| Nyquestro | Rust, Atomics, Lock-Free |
Ultra-low-latency limit-order-book engine with lock-free price levels and correctness-tested matching logic |
| Vynapse | Rust, Autodiff, NEAT, SGD, Graph IR |
Hybrid learning engine combining DEAP/NEAT-style evolution with PyTorch/TensorFlow-style autodiff and graph execution in safe Rust |
| Debate CLI | Python, Asyncio, AI APIs, TOML |
Multi-LLM debate orchestrator with recursive summarization, async provider integration, and research-grade analytics |
- 🧪 Writing a research-style whitepaper on kernel fusion design (Xyntra)
- 📈 Benchmarking latency + throughput in lock-free engines (Nyquestro)
- ✍️ Publishing research on the math behind tensor algebra and kernel fusion — covering rewrite rules, DAG transformations, and performance implications
- 🤖 Building a framework for orchestrating multi-model AI debates to surface emergent consensus and divergent perspectives
- 🖼️ Exploring local-first application architectures with embedded ML models for privacy-preserving semantic search and content organization
- 💹 Designing production-grade trading infrastructure with deterministic replay, pre-trade risk controls, and sub-microsecond latency requirements