Riley
Zurrin

Senior Data Engineer

Dakota Analytics · Data Engineer, 2024–25

• Lead a team of five engineers delivering production data platforms for enterprise clients.
• Built a greenfield platform on Dagster, DuckDB, and dbt that moves billions of records.
• Engineered real-time ingestion with Azure Event Hubs and Spark Streaming on Databricks.
• Deployed an LLM pipeline in Databricks that pulls structured JSON out of unstructured documents through a real-time inference endpoint, saving 20+ hours of manual entry every week.
• Shipped the company’s first client-facing web application, Django on Azure.
• Architected end-to-end warehouses on Databricks, Fabric, and SQL Server.
• Automated data lineage documentation with GitHub Actions and the Claude API, cutting manual upkeep ~90%.

Co-founder

Syntropic · Bert Software

Warehouse writeback without the fragile spreadsheet pipeline.

• Built and launched a revenue-generating SaaS product: a FastAPI and React spreadsheet UI that lets business users query data warehouses. It now serves live production traffic for an enterprise customer.
• Wrote an MCP server and app that embed the Syntropic UI natively inside Claude and ChatGPT.
• Orchestrated 3-stage CI/CD across Terraform-managed AWS infrastructure via Docker.
• Implemented secure connectivity to Databricks, Snowflake, and S3.
• Built the enterprise surface: RBAC, monitored webhooks, data quality checks, and approvals.
• Also building Agent Forum, a coordination layer where autonomous agents share workflows alongside humans.

Neuroscience Researcher

BC Children’s Hospital

• Analysed high-dimensional fMRI data (56 subjects, 228k+ voxels across 192 timepoints) in Python and MATLAB.
• Applied constrained PCA and SVD to tie brain activation patterns to task timing across conditions.
• First author on one paper, co-author on three more; 40+ citations combined.

Astrophysics Research Assistant

Balloon Astrophysics Group, University of Toronto

• Processed high-rate telemetry from SuperBIT, a NASA stratospheric balloon telescope flying at ~35 km.
• Assembled flight hardware at NASA’s facility in Palestine, Texas: solar panels, thermistor instrumentation, vacuum-chamber-qualified electronics.
• Evaluated image compression algorithms to squeeze more science through the downlink.

Particle Physics Research Assistant

University of Toronto

• Analysed Large Hadron Collider data with Bayesian nested sampling (SMEFiT) to constrain the masses of 21 single-particle extensions to the Standard Model.

Instructor, Data Analysis in Physics

University of Toronto & University of British Columbia

• Taught data analysis to 180+ students across four lab courses and kept the lab computing environment running.