Enterprise data entities onboarded through Azure ingestion frameworks to Databricks and PySpark.
Data Engineering Manager / Toronto, Canada
I design cloud data systems with production gravity.
11+ years across banking, telecom, and energy, turning migrations, governance, analytics platforms, and AI-assisted delivery into systems teams can actually run.
100+
entities
200M
daily records
Daily records handled through GCP data pipelines in telecom production environments.
Legacy analytics workflows migrated from scripts, SQL procedures, and Alteryx to Azure Databricks.
Developers led through a Splunk-to-Looker analytics migration covering dashboards, alerts, and data models.
What I Bring
Delivery leadership with real platform depth.
My work sits where architecture, delivery, and stakeholder alignment meet: define the ingestion pattern, coordinate source teams and product owners, harden distributed pipelines, and keep production quality visible.
I use GitHub Copilot and LLM tooling where they help the work: validation, documentation, implementation planning, and delivery accelerators that reduce ambiguity and catch mistakes earlier.
Case Study
Banking ingestion modernization on Azure.
Problem
A large banking program needed to onboard 100+ enterprise data entities through a governed ingestion path while coordinating source teams, architects, product owners, and delivery dependencies.
Approach
Led the ingestion team using Azure Data Factory, Databricks, and PySpark. Standardized configuration patterns, delivery planning, validation checks, and simulation workflows so complex onboarding could move with less rework.
Outcome
Established a repeatable delivery model for enterprise-scale ingestion, improving visibility across stakeholders and giving the team stronger guardrails for quality, documentation, and implementation speed.
How I Work
Make the platform understandable before making it bigger.
Turn migrations into operating models
I care about the repeatable path: ingestion standards, ownership, validation, monitoring, and handover. The cloud move only counts when teams can run it.
Design for the people debugging it later
Distributed pipelines need clear models, traceable logic, and failure modes people can explain under pressure. Elegance is useful only when production is loud.
Move dashboards with their meaning intact
Analytics migration is not a screenshot exercise. I focus on semantic models, query behavior, alert logic, and stakeholder trust in the numbers.
Use AI as delivery leverage
I use LLMs and Copilot for checks, drafts, test thinking, and repetitive planning work, then keep human review close to architecture and production risk.
Interactive Builds
Dashboards, media tools, and small products with a pulse.
I like projects where data has to become something people can explore, publish, explain, or act on. The sweet spot is useful software with a bit of visual rhythm.
Podcast League compares shows like data products.
Benchmark lab, podcast scoreboard, NLP synthesis, quotes library, corpus intelligence, and backtested market signals in one live interface.
View Podcast League
Benchmark / NLP / Quotes
230 posts
92 following
44 followers
Open Instagram channel
Cronisphere is the publishing layer around the research.
A real editorial experiment connecting finance, AI, markets, and podcast-driven insight into a visual social format.
View Cronisphere
Instagram / editorial system
Have a raw dataset, messy workflow, or half-formed product idea?
Pitch me the messy versionSelected Personal Builds
Proof that the hands still touch the keyboard.
Loading projects from the portfolio spreadsheet...
Get in touch