Function: Data Science Job: Machine Learning Engineer Position: Senior Immediate manager (N+1 Job title and name): AI Manager Additional reporting line to: Global VP Engineering Position location: Mumbai, Pune, Bangalore, Hyderabad, Noida. 1. Purpose of the Job – A simple statement to identify clearly the objective of the job. The Senior Machine Learning Engineer is responsible for designing, implementing, and deploying scalable and efficient machine learning algorithms to solve complex business problems. The Machine Learning Engineer is also responsible of the lifecycle of models, once deployed in production environments, through monitoring performance and model evolution. The position is highly technical and requires an ability to collaborate with multiple technical and non-technical profiles (data scientists, data engineers, data analysts, product owners, business experts), and actively take part in a large data science community. 2. Organization chart – Indicate schematically the position of the job within the organization. It is sufficient to indicate one hierarchical level above (including possible functional boss) and, if applicable, one below the position. In the horizontal direction, the other jobs reporting to the same superior should be indicated. A Machine Learning Engineer reports to the AI Manager who reports to the Global VP Engineering. 3. Key Responsibilities and Expected Deliverables– This details what actually needs to be done; the duties and expected outcomes. Managing the lifecycle of machine learning models Develop and implement machine learning models to solve complex business problems. Ensure that models are accurate, efficient, reliable, and scalable. Deploy machine learning models to production environments, ensuring that models are integrated with software systems. Monitor machine learning models in production, ensuring that models are performing as expected and that any errors or performance issues are identified and resolved quickly. Maintain machine learning models over time. This includes updating models as new data becomes available, retraining models to improve performance, and retiring models that are no longer effective. Develop and implement policies and procedures for ensuring the ethical and responsible use of machine learning models. This includes addressing issues related to bias, fairness, transparency, and accountability. Development of data science assets Identify cross use cases data science needs that could be mutualised in a reusable piece of code. Design, contribute and participate in the implementation of python libraries answering a data science transversal need that can be reused in several projects. Maintain existing data science assets (timeseries forecasting asset, model monitoring asset) Create documentation and knowledge base on data science assets to ensure a good understanding from users. Participate to asset demos to showcase new features to users. Be an active member of the Sodexo Data Science Community Participate to the definition and maintenance of engineering standards and set of good practices around machine learning. Participate to data science team meeting and regularly share knowledge, ask questions, and learn from others. Mentor and guide junior machine learning engineers and data scientists. Participate to internal or external relevant conferences and meet ups. Continuous Improvements Stay up to date with the latest developments in the field: read research papers, attend conferences, and participate in trainings to expand their knowledge and skills. Identify and evaluate new technologies and tools that can improve the efficiency and effectiveness of machine learning projects. Propose and implement optimizations for current machine learning workflows and systems. Proactively identify areas of improvement within the pipelines. Make sure that created code is compliant with our set of engineering standards. Collaboration with other data experts (Data Engineers, Platform Engineers, and Data Analysts) Participate to pull requests reviews coming from other team members. Ask for review and comments when submitting their own work. Actively participate to the day-to-day life of the project (Agile rituals), the data science team (DS meeting) and the rest of the Global Engineering team 4. Education & Experience – Indicate the skills, knowledge and experience that the job holder should require to conduct the role effectively Engineering Master’s degree or PhD in Data Science, Statistics, Mathematics, or related fields 5 years+ experience in a Data Scientist / Machine Learning Engineer role into large corporate organizations Experience of working with ML models in a cloud ecosystem Statistics & Machine Learning Statistics : Strong understanding of statistical analysis and modelling techniques (e.g., regression analysis, hypothesis testing, time series analysis) Classical ML : Very strong knowledge in classical ML algorithms for regression & classification, supervised and unsupervised machine learning, both theoretical and practical (e.g. using scikit-learn, xgboost) ML niche: Expertise in at least one of the following ML specialisations: Timeseries forecasting / Natural Language Processing / Computer Vision Deep Learning: Good knowledge of Deep Learning fundamentals (CNN, RNN, transformer architecture, attention mechanism, …) and one of the deep learning frameworks (pytorch, tensorflow, keras) Generative AI: Good understanding of Generative AI specificities and previous experience in working with Large Language Models is a plus (e.g. with openai, langchain) MLOps Model strategy : Expertise in designing, implementing, and testing machine learning strategies. Model integration : Very strong skills in integrating a machine learning algorithm in a data science application in production. Model performance: Deep understanding of model performance evaluation metrics and existing libraries (e.g., scikit-learn, evidently) Model deployment: Experience in deploying and managing machine learning models in production either using specific cloud platform, model serving frameworks, or containerization. Model monitoring : Experience with model performance monitoring tools is a plus (Grafana, Prometheus) Software Engineering Python: Very strong coding skills in Python including modularity, OOP, data & config manipulation frameworks (e.g., pandas, pydantic) etc. Python ecosystem: Strong knowledge of tooling in Python ecosystem such as dependency management tooling (venv, poetry), documentation frameworks (e.g. sphinx, mkdocs, jupyter-book), testing frameworks (unittest, pytest) Software engineering practices: Experience in putting in place good software engineering practices such as design patterns, testing (unit, integration), clean code, code formatting etc. Debugging : Ability to troubleshoot and debug issues within machine learning pipelines Data Science Experimentation and Analytics Data Visualization : Knowledge of data visualization tools such as plotly, seaborn, matplotlib, etc. to visualise, interpret and communicate the results of machine learning models to stakeholders. Basic knowledge of PowerBI is a plus Data Cleaning : Experience with data cleaning and preprocessing techniques such as feature scaling, dimensionality reduction, and outlier detection (e.g. with pandas, scikit-learn). Data Science Experiments : Understanding of experimental design and A/B testing methodologies Data Processing: Databricks/Spark : Basic knowledge of PySpark for big data processing Databases : Basic knowledge of SQL to query data in internal systems Data Formats : Familiarity with different data storage formats such as Parquet and Delta DevOps Azure DevOps : Experience using a DevOps platform such as Azure DevOps for using Boards, Repositories, Pipelines Git: Experience working with code versioning (git), branch strategies, and collaborative work with pull requests. Proficient with the most basic git commands. CI / CD : Experience in implementing/maintaining pipelines for continuous integration (including execution of testing strategy) and continuous deployment is preferable. Cloud Platform: Azure Cloud : Previous experience with services like Azure Machine Learning Services and/or Azure Databricks on Azure is preferable. Soft skills Strong analytical and problem-solving skills, with attention to detail Excellent verbal and written communication and pedagogical skills with technical and non-technical teams Excellent teamwork and collaboration skills Adaptability and reactivity to new technologies, tools, and techniques Fluent in English 5. Competencies – Indicate which of the Sodexo core competencies and any professional competencies that the role requires Communication & Collaboration Adaptability & Agility Analytical & technical skills Innovation & Change Rigorous Problem Solving & Troubleshooting Show more Show less
