Job Overview

We’re hiring a Full-Stack Data Scientist who can own the full ML lifecycle—from framing ambiguous business problems, to building statistically sound models, to shipping them into production with engineering rigor, and then operating them reliably over time.

This is not a “notebook-only” role. You will build models that create measurable business value and engineer them into durable, observable services and pipelines.

What you’ll do

1) Translate business problems into ML solutions that move metrics

• Partner with stakeholders and SMEs to understand the domain, convert real problems into analytical form, and select the right methodology (ML, statistics, optimization, simulation).
• Define success metrics, evaluation approaches, and validation plans (including baseline comparisons and monitoring strategy).

2) Build high-quality ML models (the “real data science” part)

• Design, develop, and iterate on models (forecasting, regression, classification, clustering, anomaly detection, etc.) with strong feature engineering and disciplined experimentation.
• Deliver clear, decision-ready insights and communicate methods/results to technical and non-technical audiences.

3) Engineer models into production (the “ML Engineer” part)

• Productionize prototypes into robust ML systems with appropriate error handling, versioning, reproducibility, and deployment patterns.
• Build and maintain automated pipelines for training/validation/deployment, with CI/CD practices designed for ML workflows.
• Use AWS (SageMaker) and Databricks to operationalize training and inference workflows, with a clean separation of data engineering, feature engineering, and model logic.

4) Own model lifecycle management (tracking, registry, governance)

• Track experiments and manage model artifacts with MLflow, operating a disciplined model promotion process (e.g., staging to production).
• Leverage a model registry as a centralized system for model lineage/versioning and lifecycle management.

5) Operate production ML (monitoring, alerts, and continuous improvement)

• Implement observability across model and data health: drift detection, performance regression, and actionable alerts with runbooks.
• Support and enhance existing production models (new features, improvements, reliability hardening), driving continuous improvement post-deployment.

Responsibilities:

• Demonstrated hands-on experience building ML models and deploying/operating them in production (end-to-end ownership).
• Strong Python skills; ability to write clean, testable, maintainable code (refactoring, modularity, code review discipline).
• Experience with distributed data/ML workloads in PySpark and strong SQL/data wrangling capability.
• Practical experience with AWS, especially SageMaker, and experience delivering ML workloads on Databricks.
• Experience with MLflow for experiment tracking and model lifecycle workflows.
• Strong communication skills and the ability to collaborate across functions to embed analytics into business processes.

Candidate Desired Profile:

• Experience implementing CI/CD for ML systems (tests, data/contract checks, packaging, automated deployments).
• Experience with model monitoring/drift tooling and defining retraining triggers tied to business SLAs.
• Experience with modern ML frameworks (e.g., PyTorch/TensorFlow) and GenAI/LLM workflows.
• Manufacturing/industrial analytics exposure (quality, supply chain, pricing, forecasting).

Tech stack (current/target)

• Compute/Platforms: AWS (SageMaker), Databricks
• Languages: Python, PySpark, SQL
• MLOps/Lifecycle: MLflow tracking + model registry
• Engineering practices: CI/CD, code quality, monitoring/alerting

Competencies