Waqas Bin Hamed

Currently building data pipelines and ML systems at The Home Depot. Computational math background from Waterloo — I care about how algorithms work, not just that they do.

Experience

Data Engineer June 2024 – Present
The Home Depot, Toronto, ON

  • Redesigned 150+ SAP-to-GCP pipeline triggers from scheduled polling to event-driven execution (Pub/Sub → Composer), achieving near-real-time data availability for ML workloads and eliminating timing-gap failures; enforced CI/CD checks for schema compliance before deployment.
  • Migrated 100+ Airflow pipelines to GCP Composer 3.x across two upgrade phases with modular refactoring and zero service disruption.
  • Designed dimensional models and partitioning/clustering strategies for core BigQuery tables serving 180+ stores and 30,000+ associates, improving query performance and reducing compute costs.
  • Provisioned GCP infrastructure (BigQuery, GCS, Composer) as version-controlled Terraform code, standardizing deployments and enforcing data governance across environments.

Data Scientist (Contract) February 2024 – May 2024
Hyly.AI, Ashburn, VA

  • Built a dbt-based data observability framework in BigQuery, implementing automated anomaly detection and pipeline health monitoring across multi-tenant data products and reducing downstream errors by 30–40%.
  • Developed ELT pipelines and analytics views for a real estate marketing platform, integrating 5+ data sources per client via REST APIs and enabling self-serve reporting for non-technical stakeholders.

Graduate Researcher September 2022 – June 2023
University of Waterloo, Waterloo, ON

  • Designed and implemented optimization algorithms for sum-of-norms regularized NMF, leading all algorithm development, ablation studies, and statistical benchmarking; published in Neural Computation (2026).

Data Scientist (Co-op) May 2022 – August 2022
SYMX.AI (formerly Symboticware), Sudbury, ON

  • Fine-tuned a DistilBERT fault classifier on mining maintenance logs (~500 → ~5,000 training pairs via NLP augmentation), improving classification accuracy from ~50% to ~75%; shipped as a commercial product.
  • Engineered an AWS pipeline stress-testing framework (Lambda, S3) to simulate 100s of GBs of IoT sensor data, validating ingestion capacity before production scale-up.

Data Scientist July 2020 – August 2021
Big Byte Insights, New York, NY

  • Built NLP ingestion pipelines processing 10+ GBs/day of SEC filings, Reddit posts, and financial news via REST API integrations for intellistocks.ai.
  • Applied time series forecasting and DBSCAN geospatial clustering on real estate data across ~15 homebuilder companies to generate investment reports for hedge fund clients.
  • Conducted competitive intelligence analysis via API pattern analysis and feature engineering across product SKUs to enable cross-city pricing comparisons.

Technical Skills

ML & AI: PyTorch, HuggingFace Transformers, scikit-learn, XGBoost, LangChain, Gemini API, MLflow
Data & Cloud: GCP (BigQuery, Composer, Pub/Sub, Vertex AI), AWS (S3, Lambda, EC2); Airflow, dbt, Terraform, Dataform
Languages & Tools: Python, SQL, Bash; FastAPI, Docker, Looker, Git

Education

Master of Mathematics, Computational Mathematics 2023
University of Waterloo, Waterloo, ON

  • International Master’s Award of Excellence (IMAE); Outstanding Teaching Assistant Award

Bachelor of Science, Environmental Engineering 2020
Middle East Technical University, Ankara, Turkey

  • Ranked 2nd in Department; Dean’s Honor Roll Award

Publication

Ang, A., Hamed, W., & De Sterck, H. (2026). Sum-of-norms regularized Nonnegative Matrix Factorization. Neural Computation. doi.org/10.1162/NECO.a.1482 PDF