Rayaq Siddiqui

Waterloo, ON

Graduated with Distinction. Coursework in Algorithms, Data Structures, Operating Systems, Concurrency, Compilers, Databases, Adv. C++, Computer Vision. Member of the UW Data Science Club and WATonomous.

Visit Google →

Sunnyvale, CA

C++, Rust, TypeScript, VCS, Concurrency, Microservices, Software Design & Architecture

• Building JJ, on track to become the primary VCS for every engineer at Google; contributing to its open-source project and internal Rust CLI
• Scaling JJ's distributed C++ and Rust APIs to absorb Google-wide load from engineers and coding agents alike, via concurrency, load balancing, caching, and rate limiting
• Built agent-driven development workflows – automated code review agents, scheduled agent jobs, and prompt-engineered orchestration pipelines – to accelerate the software development life cycle
• Maintaining high test coverage across JJ's Rust CLI and internal contributions, prioritized by module criticality
• Authoring design docs and technical plans for new JJ features across Google's internal developer tooling

Visit Google →

Toronto, ON

Go, GCP, Concurrency, Distributed Computing, Databases, API, Microservices

• Built on Remote Build Execution, accelerating remote builds for clients including Chrome, Android & TensorFlow
• Operated across globally distributed GCP infrastructure and sharded Spanner databases under a microservice architecture
• Developed an instance-monitoring system that kept live migrations consistent with expected system state

Visit d-Matrix →

Toronto, ON

C++, PyTorch, LLVM, MLIR, torch-MLIR

• Accelerated generative AI (LLM, diffusion) inference by 20x via systems-level architecture work in C++ on the ML compiler stack
• Implemented 2D convolution, pooling, ResNet, Stable Diffusion, LLaVA & Vision Transformer support in the frontend compiler – the first working convolution ops on the platform
• Built LLaMA 2 operations and transformations on top of the LLVM MLIR and torch-MLIR project architecture
• Led integration of the LLaMA 3 decoder and full model, achieving a 17.5x runtime speedup

Visit IBM →

Toronto, ON

Python, C++, PyTorch, Computer Vision, CUDA, AWS SageMaker

• Built scalable ML software to automate object-detection tasks in Python and PyTorch
• Pruned & quantized Facebook Research's Faster R-CNN for a facial-analysis model serving 9M users/year, cutting manual verification hours 75% and saving $35M annually
• Trained on CUDA GPUs to 99% accuracy and a 0.99 F1-score, while shrinking model size by 78%

Visit BlackBerry Limited →

Waterloo, ON

Python, TensorFlow, NLP, Elasticsearch, Docker, AWS S3, SageMaker, EC2

• Built a log anomaly detection platform combining NLP, data pipelines, and Elasticsearch in Python
• Raised NLP model accuracy from 60% to 91% and F1-score from 0.30 to 0.87
• Productionized two anomaly detection models (Autoencoder + Isolation Forest; Transformer per Google's paper) with a retraining pipeline on TensorFlow, Docker, AWS S3, SageMaker & EC2, shipped through GitLab CI/CD
• Made inference ~382x faster and cut low-level memory use 52% through memory tracing, profiling, and multiprocessing

Visit RBC →

Toronto, ON

Python, Django, SQL, Exchangelib

• Built a Python/Django dashboard integrating 7 data sources into a single operational view
• Built an automated mailing-response system that saved an estimated 600 hours/month
• Integrated a cloud database using Exchangelib, automated scripts, Django models & SQL queries

Toronto, ON

Python, JavaScript/TypeScript, React, jQuery, Node.js, Selenium

• Contributed on the Creative Pod, developing an interactive iframe with Python, JavaScript/TypeScript, React, jQuery, Node.js, and Selenium — building features, fixing bugs, and writing unit tests in a test-driven, agile environment
• Transitioned the codebase from Sinon/Chai to Jest using the Jest-Extended library, resulting in 2x faster tests running independently in parallel across threads
• Replaced library implementations with hand-written algorithms across two repositories, speeding up selected components by 2–8x

• Leading extensive research into advanced distributed ML algorithms (data parallelism, federated learning)
• Implementing a P2P architecture to enable the connection of devices across large-scale distributed networks

• A wearable that reads text in front of the user, combining optical character recognition with natural-language processing

• Winner at the Pfizer/McGill-sponsored PharmaHacks hackathon, classifying multiple diseases from gastrointestinal data

• Semantic segmentation and classification of ultrasound images with UNet and ENet to identify malignant cell clusters

• Led a campaign that raised over $80,000 across two years for pediatric cancer care and research in Northern Ontario

• Held student-government leadership across three academic years, and organized a mental-health conference for 200+ students transitioning into high school