Machine Learning Scientist

You will be responsible for building ML solutions for decision-making problems, such as planning, sequencing, routing, allocation, and resource utilization. Your key responsibilities will include: - Prototyping quickly using agentic coding tools like Claude Code-style workflows, generating scaffolds, refactoring, writing tests, and iterating on experiments while maintaining strong engineering discipline. - Developing and evaluating models in areas such as Optimization & solvers, Deep RL / Decision Intelligence, and Predictive ML. - Designing robust evaluation harnesses for offline simulation, counterfactual testing, ablations, and scenario analysis, defining KPIs and acceptance thresholds. - Collaborating with ML engineers to support productionization by considering latency/throughput constraints, monitoring, reproducibility, model versioning, and safe rollout. - Writing clear technical documentation and effectively communicating findings to both technical and non-technical stakeholders. As for the qualifications required, we are looking for someone who meets the following criteria: - 0-5 years of experience in applied ML / data science / applied research, with internships, thesis work, and strong project portfolios all being considered. - Demonstrated experience using agentic coding assistants like Claude Code or similar tools to accelerate iteration without compromising code quality. - Strong Python skills and comfort with ML tooling, with PyTorch preferred and TensorFlow also acceptable. - Solid foundations in algorithms, probability/statistics, and experimental design. - Ability to translate complex real-world problems into clear formulations with measurable success metrics. In addition to the required qualifications, the following skills and experiences are considered a strong plus or preferred: - Prior work in Deep RL methods such as PPO/SAC/DQN, offline RL, imitation learning, and MCTS/planning hybrids. - Experience with simulation-based evaluation or digital twins, including building environments/simulators, reward design, stability/debugging, and evaluation. - Familiarity with MLOps basics such as MLflow, Docker, CI/CD, and model monitoring. - Domain exposure to logistics/supply chain/industrial operations is a nice-to-have but not required. The indicative tools and technologies you will be working with include Python, PyTorch, OR-Tools/solver stacks, RL libraries like Ray RLlib/Stable Baselines, SQL, Docker, Git, and MLflow. Experience with cloud platforms is a plus. Please note that no additional details about the company were included in the job description. You will be responsible for building ML solutions for decision-making problems, such as planning, sequencing, routing, allocation, and resource utilization. Your key responsibilities will include: - Prototyping quickly using agentic coding tools like Claude Code-style workflows, generating scaffolds, refactoring, writing tests, and iterating on experiments while maintaining strong engineering discipline. - Developing and evaluating models in areas such as Optimization & solvers, Deep RL / Decision Intelligence, and Predictive ML. - Designing robust evaluation harnesses for offline simulation, counterfactual testing, ablations, and scenario analysis, defining KPIs and acceptance thresholds. - Collaborating with ML engineers to support productionization by considering latency/throughput constraints, monitoring, reproducibility, model versioning, and safe rollout. - Writing clear technical documentation and effectively communicating findings to both technical and non-technical stakeholders. As for the qualifications required, we are looking for someone who meets the following criteria: - 0-5 years of experience in applied ML / data science / applied research, with internships, thesis work, and strong project portfolios all being considered. - Demonstrated experience using agentic coding assistants like Claude Code or similar tools to accelerate iteration without compromising code quality. - Strong Python skills and comfort with ML tooling, with PyTorch preferred and TensorFlow also acceptable. - Solid foundations in algorithms, probability/statistics, and experimental design. - Ability to translate complex real-world problems into clear formulations with measurable success metrics. In addition to the required qualifications, the following skills and experiences are considered a strong plus or preferred: - Prior work in Deep RL methods such as PPO/SAC/DQN, offline RL, imitation learning, and MCTS/planning hybrids. - Experience with simulation-based evaluation or digital twins, including building environments/simulators, reward design, stability/debugging, and evaluation. - Familiarity with MLOps basics such as MLflow, Docker, CI/CD, and model monitoring. - Domain exposure to logistics/supply chain/industrial operations is a nice-to-have but not required. The indicative tools and technologies you will be working with include Python,