Machine Learning Engineer

Phantasma Labs; Berlin; Hybrid; EUR 90,000 - 150,000; Full time

Benefits offered by this job

Ownership from day 1

Strong team collaboration

Flexible working hours

Hybrid setup in Berlin Mitte

Company laptop

Job summary

Phantasma Labs in Berlin is hiring a Machine Learning Engineer to advance our RL agents and Digital Twin environments for production scheduling. You will own core AI, shaping models that optimize schedules in real factories and collaborate closely with our engineering team and customers.

Join a small international team with a flexible hybrid setup in Berlin, contributing to production-ready AI that scales across manufacturers worldwide.

Qualifications

• Background in Computer Engineering, Mathematics, Machine Learning or related field.
• 5+ years of Python experience and 2+ years in factory shopfloor operations or 2+ years experience with ERP/MES systems.
• Strong understanding of manufacturing processes across production environments.
• Basic understanding of reinforcement learning and RL algorithms.

Responsibilities

• Write robust, scalable production-ready Python code.
• Provide code reviews and mentorship to maintain high coding standards.
• Write unit and integration tests using unittest and pytest.
• Design, engineer and optimize features in the Digital Twin for RL simulations.
• Create, optimize and maintain training and evaluation scripts for RL agents.
• Set up and maintain Python environments using modern tools (uv and conda).
• Work with git (GitHub) and collaborate across teams.
• Participate in customer calls to translate requirements into technical features.
• Brainstorm ideas to improve RL agents, including algorithms and reward functions.

Skills

Python

Reinforcement learning

Factory knowledge

ERP/MES familiarity

CI/CD

PyTorch

GitHub

Education

Computer Engineering / Mathematics / ML

Tools

PyTorch

NumPy

Pandas

GitHub Actions

Job description

At Phantasma Labs, we're building AI-powered production scheduling software for manufacturers. Production scheduling is a complex problem: priorities change, machines go down, new orders come in and planners constantly need to adapt. Many factories still rely heavily on manual planning to manage this complexity. We approach this differently. Our technology combines reinforcement learning with Digital Twin environments to train AI agents in simulation, allowing manufacturers to create and optimize production schedules without relying on large historical datasets. We're already working with manufacturers across Europe, the US and Japan and partner with ERP and MES providers to bring our technology into real production environments. We're a small, international team based in Berlin with a flexible remote setup, and there's still a lot to build as we continue developing the product and bringing it to more manufacturers.

Tasks

As a Machine Learning Engineer, you'll work on the core AI behind our production scheduling software. You'll develop and improve our reinforcement learning agents, work on the Digital Twin environments they train in and help turn complex manufacturing problems into models that work in real factories. You'll have a lot of ownership over how we approach these problems and work closely with both our engineering team and customers. We're looking for someone who brings strong Python experience together with an understanding of manufacturing environments and wants to apply reinforcement learning to real‑world production challenges.

• Write robust, scalable and production-ready Python code
• Provide code reviews, guidance, and mentorship to fellow developers to maintain high coding standards
• Write unit and integration tests using tools such as unittest and pytest
• Design, engineer and optimize features in the Digital Twin for reinforcement learning simulations using Python, NumPy and Pandas
• Create, optimize and maintain training and evaluation scripts for RL agents
• Set up and maintain Python environments using modern tools such as uv and conda
• Work collaboratively using git (GitHub)
• Participate in customer calls to understand requirements and translate them into actionable technical features
• Brainstorm and develop ideas to improve the RL agent, including algorithms, reward functions and architecture

Requirements Must haves

• A background in Computer Engineering, Mathematics, Machine Learning, Industrial Engineering or a related field
• 5+ years of experience with Python 2+ years of experience in factory shopfloor operations as an engineer or planner, or 2+ years of experience working with ERP/MES systems for factories
• Strong understanding of manufacturing processes across different production environments, including discrete manufacturing, line production and engineer-to-order
• Basic understanding of reinforcement learning and experience developing or applying RL algorithms

Nice to haves

• 3+ years of experience in factory shopfloor operations as an engineer or planner, or 3+ years of experience working with ERP/MES systems for factories
• Research experience in developing RL algorithms
• Experience with CI/CD pipelines, particularly GitHub Actions
• Experience with libraries and tools such as PyTorch, Optuna and MLflow

Benefits

• Ownership from day 1: Work in a small team with fast feedback and see the impact of what you build
• Collaborate with a strong team: Work alongside ML specialists developing our AI optimization technology and software engineers building the production-grade systems around it
• A supportive, open culture: clear communication, strong collaboration and flat hierarchies
• Flexible working hours & hybrid setup: work remotely or from our Co-working space in Berlin Mitte, whatever helps you do your best work
• Company laptop: We'll provide the equipment you need to do your work

Sounds Like a Fit?

If this sounds like your type of challenge, we'd love to hear from you! Don't worry if you don't tick every single box. If you're