ML Intern

Sowiz; United States; Remote; USD 70,000 - 110,000; Full time

Job summary

Sowiz is seeking a recent graduate to join our machine learning team focused on forecasting and decision-support solutions. You will design, train, and iterate models leveraging tree-based methods (CatBoost, XGBoost, LightGBM) and neural networks, while building scalable data pipelines using SQL and NoSQL.

Candidates should demonstrate advanced Python skills, experience with NumPy/Pandas, and a strong math foundation.

Qualifications

• Recent graduate with a quantitative degree (CS/Math/Statistics/EE).
• Advanced production-level Python programming required.
• Hands-on experience with tree-based models (CatBoost, XGBoost, LightGBM).
• Solid knowledge of SQL and NoSQL databases.
• Strong foundation in linear algebra, calculus, probability, and optimization.

Responsibilities

• Develop, train, and iterate ML models for forecasting and decision-support.
• Architect and optimize models using tree-based methods and neural networks.
• Build and maintain data pipelines with SQL/NoSQL data stores.
• Tune hyperparameters, apply model compression, and maximize inference efficiency.
• Design validation strategies and track experiments with MLflow or Weights & Biases.

Skills

Python programming

NumPy

Pandas

Vectorized code

Neural networks

SQL

NoSQL

Linear algebra

Multivariable calculus

Probability

Education

B.Tech / M.Tech / MS in CS/Statistics/Math/EE

Tools

CatBoost

XGBoost

LightGBM

PyTorch

TensorFlow

Job description

Role & responsibilities

• Model Development & Iteration: Develop, train, and rigorously iterate on machine learning models tailored for forecasting and decision-support use cases.
• Modeling Expertise: Architect and optimize robust models utilizing strong expertise in tree-based models (CatBoost, XGBoost, LightGBM) alongside deep learning frameworks (PyTorch/TensorFlow).
• Data Engineering & Storage: Extract, transform, and manage large-scale datasets leveraging both SQL and NoSQL databases to build high-throughput feature engineering pipelines.
• Optimization & Deployment: Apply hyperparameter tuning, model compression, and optimization techniques to maximize inference efficiency.
• Evaluation & Experimentation: Design rigorous validation strategies, analyze error distributions, and track experiments using tools like MLflow or Weights & Biases.

Preferred candidate profile

• Education: Recent Graduate (holding a B.Tech, M.Tech, MS) in Computer Science, Statistics, Mathematics, Electrical Engineering, or a closely related quantitative field.
• Programming Proficiency (Required): Advanced, production-level Python programming experience is strictly required, accompanied by deep familiarity with NumPy, Pandas, and writing efficient, vectorized code.
• Modeling Expertise: Strong, hands‑on understanding and implementation experience with tree-based models (CatBoost, XGBoost, LightGBM) as well as neural networks.
• Data Systems: Solid working knowledge of querying and managing data across both SQL and NoSQL databases.
• Core Math Foundation: Strong grasp of linear algebra, multivariable calculus, probability, and optimization theory.