Lead Data Scientist

Location: Vancouver

Experience: 4 - 8 Years

Business Unit: DirectCore

Job Type: Full Time

Openings: 1

We are seeking a highly skilled Lead Data Scientist with a strong foundation in AI/ML engineering to join our dynamic team in Vancouver. The ideal candidate will possess a robust understanding of data science principles, AI-assisted workflows, and engineering practices. This role requires a hands-on approach to solving complex business problems through data-driven insights and the development of reliable analytical assets.

A. Computer-use and engineering fluency Every hire must be able to operate as a modern technical builder, not as a notebook-only analyst. Uses Python and SQL fluently. Works in Git with branches, pull requests, code review, and reproducible environments. Comfortable with terminal, package management, notebooks, scripts, APIs, logs, and containers. Can read data from warehouses or lakehouse environments such as Snowflake, Databricks, BigQuery, Redshift, Spark, or equivalent. Can turn exploratory work into reusable functions, scripts, tests, and documented assumptions. Can troubleshoot failed jobs, broken queries, bad joins, package conflicts, and data-quality issues without immediately requiring an engineer. B. AI-native delivery workflow AI-assisted coding and analysis is a hard requirement. Uses LLMs or coding agents for exploration, code generation, refactoring, documentation, test creation, debugging, or analysis acceleration. Can explain what AI-generated output they accepted, rejected, rewrote, and tested. Can detect plausible but wrong AI output. C. Applied data science capability Focus practical data science for delivery. Candidates should be able to use data to clarify business problems, build reliable analytical assets, evaluate options, and support implementation decisions in messy client environments. Working confidently with messy enterprise data: missing values, inconsistent definitions, broken joins, sparse history, duplicated records, and changing business rules. Building practical analytical workflows in Python and SQL that can be reused by other team members. Understanding forecasting, experimentation, optimization, and ML concepts well enough to apply or evaluate them pragmatically. Knowing when a simple analytical method is sufficient and when deeper modeling support is required. Communicating findings, assumptions, data limitations, and recommended next steps in a way that delivery leads and client stakeholders can act on. D. Data and ML engineering literacy Not every hire needs to be an ML engineer, but every hire must understand production constraints. Understands batch pipelines, feature generation, data contracts, basic orchestration, model artifacts, environment management, and CI/CD concepts. Can work with data engineers and ML engineers without throwing work “over the wall.” Can create or interpret data-quality checks. Understands model versioning, data versioning, reproducibility, deployment handoff, monitoring, and rollback concepts. Can produce a model card, validation note, or handoff document that another team can operate.