Machine Learning Engineer

KSB is a leading supplier of pumps, valves and related service. Our reliable, high‑efficiency products are used in applications wherever fluids need to be transported or shut off, covering everything from building services, industry and water transport to wastewater treatment, power plant processes and mining. Founded in 1871 in Frankenthal, Germany, the company has a presence on all continents with its own sales and marketing organisations and manufacturing facilities.

Around the globe, more than 190 service centres and around 3,500 service specialists are on hand to provide local inspection, servicing, maintenance and repair services under the KSB Supreme Serv brand. Innovative technology that is the fruit of KSB’s research and development activities forms the basis for the company’s success.

People. Passion. Performance. It is these three success factors that make KSB the company it is today.

At KSB, we recognise that it is people who actually make the difference – the people we employ and the people we serve. This is why we are committed to equal rights and treatment worldwide and never lose sight of the aspects ecology and sustainability when manufacturing our products.

KSB GIW, Inc.

Department:
Engineering, Research & Development

Reports to:

Metallurgical and Materials R&D Lab Manager

Location:

Grovetown, GA, USA (onsite)

Shift: First

FLSA Status:
Salary Exempt

Our R&D group is expanding its use of machine learning to solve real engineering problems, and we’re looking for a sharp, hands‑on early‑career engineer to join the team.

You’ll work at the intersection of machine learning and the physical world to build AI systems that learn from real industrial data and connect with the engineering models behind them. The role lives where machine learning meets scientific computing: surrogate modeling, data‑driven approximations of physical systems, and ML models that respect the underlying engineering principles.

You’ll build the data foundation that powers this work, implement and train models that bridge physics‑based simulation with modern machine learning, and work closely with an experienced technical lead who will guide your growth across data engineering, scientific ML, and emerging AI tooling.

• Build and maintain the data foundation: ingestion, cleaning, transformation, validation, and metadata standards
• Implement and train machine learning models using Python and modern frameworks (PyTorch)
• Contribute to applied AI tooling that supports the broader R&D workflow
• Develop visualization and dashboard interfaces that present results to end users
• Run experiments, track results, and report findings against defined targets
• Help bring prototype code to production quality: testing, documentation, version control
• Collaborate with team members across engineering disciplines

• Education:
  
  Bachelor’s degree required; master’s preferred in Computer Science, Engineering, Applied Math, Physics, or a related field
• Experience:
  
  1–3 years of professional or substantial project experience in machine learning, data engineering, or scientific computing

• Solid Python skills with hands‑on experience using core libraries:
  • Machine learning:
    PyTorch, scikit-learn
  • Data:
    Num Py, pandas
  • Scientific computing:
    Sci Py, Matplotlib
• Foundational understanding of scientific computing: numerical methods, simulation concepts, or modeling of physical systems — this is essential to the role
• Foundational understanding of neural networks, model training, and optimization
• Experience with version control (Git) and working in a Linux environment
• Strong written and verbal communication skills
• Collaborative, coachable attitude

• Experience building and maintaining data pipelines, metadata schemas, and data quality frameworks
• Exposure to scientific / physics-informed machine learning (surrogate modeling, embedding physical constraints into ML models)
• Background in CFD, simulation, computational mechanics, or applied physics
• Familiarity with agentic AI / LLM frameworks (Lang Chain, Lang Graph, or similar) enough to collaborate effectively, not lead
• Experience with Jupyter,…