This 12-month remote contract role places you on an international machine learning project through Dijkstrack for one of our global technology partners. You’ll work within a distributed engineering team to design, train, and deploy ML models using modern frameworks like PyTorch. Dijkstrack engineers enjoy access to technical community support, structured delivery processes, and the opportunity to embed within long-term global product teams.
Key Duties & Responsibilities
- Design, train, and deploy machine learning and deep learning models using PyTorch
- Build and maintain ML pipelines , model training workflows, and data processing components
- Work with product and engineering teams to integrate models into production systems
- Analyse datasets, performance metrics, and optimise models for accuracy and efficiency
- Implement MLOps best practices for model versioning, deployment, and monitoring
- Document models, datasets, evaluations, and maintain reproducibility standards
- Participate in code reviews and uphold engineering quality in ML codebases
Essential Skills & Technical Requirements
Strong proficiency in Python for ML developmentHands‑on experience with PyTorch (TensorFlow experience also welcome)Solid understanding of machine learning fundamentals, model architectures, training loopsAbility to work with large datasets, ETL pipelines, and model optimisationFamiliarity with Git workflows, remote collaboration, and agile delivery modelsBeneficial / Nice‑to‑Have Experience
Experience with ML Ops tools like MLflow, Weights & Biases, SageMaker, or similarExposure to cloud‑based ML deployments (AWS, Azure, GCP)Knowledge of microservice integration for ML modelsPrior work in AI product teams or international ML research environmentsBenefits of Working with Dijkstrack
Join a network of engineers delivering advanced ML solutions internationallyTechnical community of senior engineers across data, backend, and product domainsRemote‑first culture, global exposure, and structured technical delivery support#J-18808-Ljbffr