Niche - Technical

Associate Data Scientist

You will be trained and developed to become our dedicated expert on our enterprise AI platform — the system that drives data-driven decision-making across multiple facilities. You won't be expected to arrive knowing everything. You will be expected to learn fast, ask great questions, and care deeply about getting the answer right. Over time, this role owns a critical function: training, evaluating, and managing the machine learning models that power our enterprise AI platform — determining which models are ready for production (champion models) and continuously improving them.

As you ramp up, your responsibilities will grow to include:

• Learning our enterprise AI platform architecture and how it connects to manufacturing data across our facility network
• Training machine learning models on industrial datasets — process sensor data, quality measurements, operational metrics, and more
• Evaluating model performance using appropriate statistical methods and determining readiness for production (champion model selection)
• Managing the model lifecycle: training, validation, deployment, monitoring, and retraining
• Working with data engineers to define data transformation and feature engineering requirements
• Building visualizations and reports that communicate model outputs and insights to operations and leadership teams
• Processing, cleansing, and verifying the integrity of data used for analysis
• Performing ad-hoc analysis on structured and unstructured datasets and presenting results clearly
• Staying current on developments in machine learning, industrial AI, and relevant platform capabilities

A core responsibility of this role is owning the process of training candidate models, evaluating their performance, and deciding which model should be promoted to 'champion' status in production. This requires both technical rigor and good judgment. You will:

• Design and run model training experiments using sound train/validation/test methodology
• Compare candidate models across relevant performance metrics (accuracy, precision/recall, RMSE, drift, etc.)
• Document evaluation rationale and maintain a clear record of model versions and decisions
• Recommend champion model promotions and flag underperforming models for retraining or replacement
• Develop an understanding of how model outputs affect real manufacturing decisions — and weigh that context in your evaluations

Required:

• Bachelor's or Master's degree in Data Science, Statistics, Computer Science, Engineering, Mathematics, or a related quantitative field
• Solid foundation in machine learning concepts: supervised/unsupervised learning, model evaluation, overfitting, cross-validation
• Programming proficiency in Python, including core data science libraries (pandas, Num Py, scikit-learn)
• Understanding of statistical fundamentals: hypothesis testing, distributions, regression, and model diagnostics
• Exposure to data visualization tools or libraries (Tableau, Power BI, Matplotlib, Seaborn, or similar)
• Strong analytical thinking and ability to approach ambiguous problems methodically
• Excellent communication skills — you can explain what a model does and why it matters to someone who isn't a data scientist
• Intellectual curiosity and eagerness to learn in an industrial environment that may be new to you
• Ability to manage your own time and work independently while collaborating within a broader team