Role: “Data & ML Platform Development Lead”

Location: Remote – Canada (Travel to Toronto as required)

Job Type: Contactor/Full Time

Interview: 2 Technical Round and 1 onsite Client round.

:

Exp. Required: 8+ Year’s

We need a hands on lead.

We're looking for a hands-on development lead for a production data-and-ML platform — someone who leads the build and has the architecture judgment to set technical direction. The core profile is a data scientist with data-engineering depth to direct the work.

The role

Lead the build of a production data and ML platform for a large enterprise in the telecom identity and fraud-analytics space, delivered under a build-operate-transfer model. The platform ingests multi-source data, turns it into governed, business-ready datasets, and serves an ensemble of machine-learning models that produce trust and fraud scores.

Lead the team that builds it, be hands-on in the code yourself, and own how the platform is designed — for correctness, reusability, and extensibility. This is a development-lead role with architecture ownership: The person is expected to set the design and build in it, giving clear direction to the data-engineering and ML developers on the team.

What the role will own

- Platform architecture across the full stack — the Lakehouse (medallion S3 / Iceberg / Glue), batch pipelines on Spark/EMR, orchestration (Step Functions, EventBridge, Lambda), and the ML layer (feature engineering, training, batch scoring, monitoring).
- The ML design that makes the platform reusable — a model contract, a feature store, a label store, versioned scoring configurations, and an ensemble framework that lets current models and features be onboarded as configuration rather than one-off code.
- MLOps rigor — the train → evaluate → register → approve → deploy gate chain, controlled model promotion, drift detection,



and a clean boundary between what the platform recommends and what the client decides.
- DataOps rigor — data-quality gates, schema governance and evolution, SLOs, backfill and reprocessing, and disaster recovery.
- Extensibility as a first-class design goal — making it cheap to onboard a new data source, new data from an existing source, a new feature, model, or use case, with the right governance and approvals at each step.
- Leading the build and knowledge transfer — setting the design, leading and reviewing the build hands-on, and transferring ownership to the client's team as part of the build-operate-transfer model.

Must-have

- Extensive data-science foundation: you've built, evaluated, and productionized ML models (ensembles, feature engineering, point-in-time correctness, model evaluation and explainability) — not just consumed them.
- Data-engineering depth to direct it: you can design and build a production batch data platform — Lakehouse patterns, Spark, orchestration, data quality, and schema governance — and lead senior data engineers.
- Working understanding — familiarity with the core data/ML services (S3, EMR, SageMaker, Glue, Step Functions, Lambda) is enough to start; deeper expertise is preferred and can be grown on the job.
- Platform thinking — Design and build for reuse and extensibility, not just for the model in front of you, and you can explain the trade-offs to both developers and client stakeholders.
- Proven ability to lead a development team and set technical direction in a client-facing delivery setting.

Preferred

- Infrastructure-as-code (Terraform) and MLOps tooling (MLflow, model/feature registries).
- Relevant AWS certification (Solutions Architect and/or Machine Learning Specialty).
- Prior experience in a build-operate-transfer or consulting delivery model.

Location & travel The role is based in Canada with travel to Toronto as required to work alongside the client and the delivery team.

📌 Data & ML Platform Development Lead (Toronto)
🏢 Centraprise
📍 Toronto