Machine Learning Engineer

CPGvision is a recognized leader in Trade Promotion Management (TPM), Trade Promotion Optimization (TPO), and Revenue Growth Management (RGM). Leveraging the power of the Salesforce platform, they enable consumer goods companies to achieve their RGM objectives through a fully integrated, user-friendly solution suite. The Data Science team builds scalable machine learning models that power advanced analytics for the CPG and Retail industries, helping global brands optimize pricing and promotional strategies.

• A degree in Computer Science, Data Science, Engineering, Applied Mathematics, or a related field.
• Minimum 2 years of commercial experience as a Machine Learning Engineer, Data Engineer, or a similar role with strong ML focus.
• Strong Python skills with emphasis on best practices (OOP, type hints, testing, clean architecture, packaging).
• Hands-on experience training and deploying gradient boosting models (LightGBM, XGBoost).
• Working knowledge of AWS (S3, EC2, Lambda) and ability to set up and manage cloud-based ML workloads.
• Proficiency in Docker for containerizing ML services and pipelines.
• Solid SQL skills for data extraction and transformation.
• Comfortable working in a Linux terminal environment.
• Communicative English sufficient for reading documentation, code reviews, and presenting technical decisions.

Nice to have:

• Experience with experiment tracking and model registry tools (e.g. MLflow).
• Experience with OCR tools.
• Experience with deep learning models for time-series, particularly Temporal Fusion Transformers (TFT) or similar.
• Experience with workflow orchestration (e.g. Airflow, Prefect, Step Functions).
• Familiarity with hyperparameter optimization frameworks (e.g. Optuna).
• Knowledge of Infrastructure as Code (e.g. Terraform, CloudFormation).
• Familiarity with CPG/FMCG or Retail data domains.
• Experience with Explainable AI tools (e.g. SHAP).

ML Pipeline Development:

• Design, build, and maintain end-to-end machine learning pipelines covering data ingestion, feature engineering, model training, evaluation, and serving.

Model Training & Optimization:

• Train and optimize gradient boosting models (LightGBM, XGBoost) and deep learning architectures such as Temporal Fusion Transformers (TFT) for time-series forecasting at scale.

Cloud Infrastructure:

• Deploy and manage ML workloads on AWS (S3, EC2, Lambda), including containerized training jobs, scheduled retraining, and model artifact management.

Production Deployment:

• Package models for production use with proper versioning, monitoring, and automated testing.
• Ensure reproducibility and traceability of experiments.

Code Quality & Best Practices:

• Write clean, well-tested, modular Python code following software engineering best practices (OOP, design patterns, typing, linting, CI/CD).

Collaboration:

• Work closely with Data Scientists and business stakeholders to translate analytical prototypes into production-ready solutions and communicate technical decisions clearly.