We are seeking a Data Scientist with strong expertise in classical machine learning and predictive analytics to develop custom models for equipment manufacturing and industrial environments. The ideal candidate will have hands-on experience building machine learning models from the ground up, selecting the most suitable algorithms based on business needs, and transforming manufacturing data into actionable insights.
This role focuses on addressing real-world industrial challenges, including equipment failure prediction, quality forecasting, demand forecasting, process optimization, predictive maintenance, and production efficiency enhancement.
Key Responsibilities
Develop predictive models using classical machine learning techniques.
Design, train, validate, and deploy machine learning models using manufacturing and operational data.
Analyze structured and time-series data from production equipment, sensors, ERP systems, MES platforms, and quality management systems.
Perform feature engineering, data exploration, and statistical analysis.
Build scalable data preprocessing, model training, and validation pipelines.
Evaluate multiple algorithms and identify the most effective solutions based on business objectives and performance metrics.
Interpret model outputs and communicate insights to engineering, operations, and business stakeholders.
Collaborate with manufacturing, process engineering, quality assurance, and maintenance teams to understand operational challenges and identify opportunities for data-driven improvements.
Deploy machine learning models into production environments and monitor model performance.
Continuously improve model accuracy, reliability, and scalability.
Required Technical Skills
Machine Learning
Strong understanding and practical experience with classical machine learning algorithms, including:
Linear Regression
Logistic Regression
Decision Trees
Random Forest
Gradient Boosting
XGBoost
LightGBM
CatBoost
Support Vector Machines (SVM)
K-Nearest Neighbors (KNN)
Naïve Bayes
Clustering Techniques (K-Means, DBSCAN, Hierarchical Clustering)
Principal Component Analysis (PCA) and Dimensionality Reduction
Time Series Forecasting (ARIMA, SARIMA, Prophet)
Ensemble Learning Methods
Anomaly Detection Techniques
YOUR PROFILE
Statistical Knowledge
Hypothesis Testing
Probability
Experimental Design
Statistical Modeling
Regression Analysis
Sampling Techniques
Confidence Intervals
Time Series Analysis
Programming
Python
SQL
Python libraries:
pandas
NumPy
scikit-learn
SciPy
statsmodels
XGBoost
LightGBM
CatBoost
Matplotlib
Seaborn
Data Engineering Skills
Data Cleaning
Feature Engineering
Data Integration
ETL Pipelines
Handling Missing Values
Data Validation
Manufacturing Domain Knowledge (Preferred) Experience with:
Industrial Manufacturing
Equipment Manufacturing
Automotive
Heavy Engineering
Process Manufacturing
Factory Automation
Understanding of:
Production KPIs
OEE
Downtime Analysis
Root Cause Analysis
MES
ERP
SCADA
PLC Data
Sensor Data
IIoT
Model Development Expectations Candidates should demonstrate the ability to:
Select appropriate algorithms based on business problems.
Build machine learning models from scratch using Python.
Engineer meaningful features from manufacturing datasets.
Optimize hyperparameters.
Evaluate models using appropriate metrics.
Explain model decisions using interpretable techniques.
Deploy models into production.
Experience with AutoML tools alone is not sufficient.
Required Qualifications
Strong problem-solving and analytical skills.
Experience working with structured industrial datasets.
Excellent communication and stakeholder management skills.
Preferred Qualifications
Experience in predictive maintenance projects.
Experience with manufacturing analytics.
Knowledge of MLOps practices.
Familiarity with Docker and cloud platforms (AWS, Azure, or GCP).
Exposure to edge analytics or IoT-based machine learning.
Experience integrating ML models into enterprise applications.
Nice to Have
Knowledge of optimization techniques.
Survival Analysis and Remaining Useful Life (RUL) modeling.
Reinforcement Learning (basic understanding).
Digital Twin concepts.
Knowledge of reliability engineering.
Experience working with streaming data.
Success Measures Within the first 6-12 months, the successful candidate should be able to:
Develop production-ready predictive models for equipment health, quality, or process optimization.
Improve prediction accuracy through feature engineering and model tuning.
Collaborate effectively with manufacturing and engineering teams to deliver measurable business outcomes.
Deploy and monitor machine learning solutions that support operational decision-making.