Job
Description
We are seeking a highly capable Data Scientist cum Machine Learning Engineer to support the design, development, and deployment of advanced AI-driven solutions for real-time defect prediction and closed-loop control in a manufacturing environment. The role involves building robust data pipelines, developing high-accuracy ML models, and integrating those models into real-time control systems for automated corrective actions. Role & responsibilities: Validate and maintain the correct flow of data from InfluxDB/CDL to Smartbox/Databricks , ensuring data consistency and readiness for modeling. Develop and support data pipelines for both batch and real-time processing, including ad-hoc data extraction and corrections. Build and deploy three separate machine learning models to predict three distinct defect types using historical time-series data, aiming to forecast outcomes for the next 5 minutes. Ensure model performance meets predefined benchmarks: Accuracy 95% Precision & Recall 80% Generate interpretable insights through feature importance analysis. Apply advanced techniques such as artificial sampling , feature engineering , and dimensionality reduction to enhance model performance and reliability. Deploy models on Smartbox edge devices and integrate outputs with manufacturing PLCs to suggest setpoint adjustments aimed at preventing defects. Design and implement an output pipeline that shares model recommendations with the correct PLC tags in real-time. Develop a robust automated retraining pipeline in the cloud to handle model drift by comparing predicted recommendations against actual production centerlines. Collaborate closely with cross-functional teams including data scientists, automation engineers, and operations personnel. Preferred candidate profile: Bachelors in Computer Science, Data Science or a related discipline. Strong expertise in machine learning , particularly time-series forecasting and classification models. Proficiency in Python and relevant ML libraries (e.g., scikit-learn, pandas, NumPy, PyTorch, or TensorFlow). Solid experience in data engineering , including ETL pipeline development and data validation. Familiarity with platforms and tools like Databricks , InfluxDB , and Smartbox . Understanding of PLC communication protocols and real-time industrial control systems. Experience with model deployment on edge devices and cloud platforms (AWS, Azure, or GCP). Working knowledge of MLOps , model monitoring , and automated retraining mechanisms .
