Computer Engineering @ Toronto Metropolitan University
Building deterministic trading infrastructure, market-data systems, and low-latency C++20 software.
Trading & Simulation
Python Event-Driven Simulation Backtesting Execution Modeling Market Microstructure Monte Carlo Methods
Systems Engineering
C++20 Low-Latency Systems Binary Logging Deterministic Replay Multithreading Lock-Free SPSC Queues
Data & Tooling
Pandas NumPy Polars PyArrow pytest ruff mypy GitHub Actions
Hardware & Architecture
VHDL Quartus II Computer Architecture RTL Design
🔹 Meridian
Deterministic, event-driven backtesting engine for intraday futures.
Python, Event-Driven Simulation, Execution Modeling, Market Microstructure
- Built for causal integrity, with strict session handling and feature generation designed to prevent look-ahead and leakage.
- Models regime-adaptive execution, including slippage, timing, and fill-time risk constraints to avoid unrealistic strategy results.
- Designed for reproducibility, with CI-gated testing, determinism checks, and artifact-backed outputs for auditable runs.
🔹 Stratos
Deterministic price-time priority matching engine with oracle diff testing, replay, and verification harnesses.
C++20, Limit Order Books, Matching Engines, Low-Latency Systems
Under construction — Q1 2026.
- Implements price-time priority matching with fixed-point prices, deterministic order flow, and fail-fast invariants.
- Includes a slow-but-correct oracle engine plus seeded differential tests to compare fills and state across randomized event streams.
- Being extended with binary replay, benchmark instrumentation, and adversarial hardening for correctness-first systems development.
🔹 Auro
C++20 market-data systems project for capture, compact binary storage, and deterministic replay of crypto exchange feeds.
C++20, Market Data, Boost.Asio/Beast, simdjson, Systems Experimentation
- Captures live exchange market-data messages and stores them in a compact append-only binary format for offline analysis and benchmarking.
- Separates network I/O from downstream processing using a lock-free queue architecture to keep ingestion paths responsive under bursty flow.
- Supports deterministic replay for parser benchmarking, systems experiments, and repeatable downstream analysis workflows.
🔹 Aegis
Real-Time EMS Navigation, Triage & Algorithmic Routing Telemetry.
React, TypeScript, FastAPI, Python, OSMnx, NetworkX
- Built as an operator-facing emergency response system combining routing, telemetry, dispatch visibility, and triage workflows.
- Added algorithm-comparison and rerouting capabilities for scenario-based route evaluation and response planning.
- Awarded York Region's Best Community Impact Hack at CTRL+HACK+DEL 2.0.
- Deterministic simulation and replay systems
- Market data capture, parsing, and execution infrastructure
- Matching engines and exchange-style correctness testing
- Latency-aware C++ systems and performance profiling
- Event-driven architecture for finance and real-time analytics
- Advancing Stratos into a hardened matching engine with replay, verification, and benchmark artifacts
- Continuing to improve Meridian around correctness, execution realism, and research workflow quality
