Role Description At Phantasma Labs, we’re building smart software that helps factories plan better. Our AI co-pilot helps factories schedule production more efficiently, adapt to changes faster, and make the most of available resources, without needing tons of data or complicated tools. We work with manufacturers across Europe and the US and team up with ERP and MES providers to bring our solution directly into real production workflows. We’re based in Berlin, but remote-first, so you don’t have to be here to join us. We’re a lean, international team with big ambitions, building what we believe is one of the most advanced production planning tools in the world. - Write robust, scalable, and production-ready Python code - Provide code reviews, guidance, and mentorship to fellow developers to maintain high coding standards - Write unit tests and integration tests (unittest, pytest, etc.) - Design, engineer, and optimize features in the digital twin for Reinforcement Learning (RL) simulations using Python (Python data structures, NumPy, Pandas, etc.) - Create, optimize, and maintain training and evaluation scripts (for RL agents) - Set up and maintain Python environments using modern tools (uv, conda, etc.) - Work collaboratively using git (GitHub) - Participate in customer calls to understand and translate requirements into actionable technical features - Brainstorm ideas to improve the RL agent, including algorithms, rewards, and architecture Qualifications - A background in Computer Engineering/Mathematics/Machine Learning/Industrial Engineering or related field - 5+ years of Python experience - 2+ years of experience in factory shopfloor operations as engineer or planner - Or alternatively, 2+ years experience in ERP/MES systems for factories - Strong grasp of manufacturing processes across various domains like discrete manufacturing, line production, and engineer-to-order, etc. - Basic understanding and experience in developing/application of RL algorithms Requirements - 3+ years of experience in factory shopfloor operations as engineer or planner - Or alternatively, 3+ years experience in ERP/MES systems for factories - Research experience in developing RL algorithms - CI/CD pipeline experience (GitHub actions) - Experience in using libraries like Pytorch, Optuna, MLflow Benefits - Ownership from day 1: small team, fast feedback, visible results - Collaborate with a strong team: work alongside highly skilled ML specialists working on cutting-edge AI optimization, and experienced software engineers building the production-grade systems that bring it to life - 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 - A company laptop to support your work Company Description At Phantasma Labs, we’re building smart software that helps factories plan faster and better. Our AI figures out how to schedule production more efficiently, adapt to changes faster, and make the most of available resources – without needing tons of data or complicated tools. We train our models in simulated factory environments (Digital Twins) using reinforcement learning, which lets us find better solutions way faster than traditional systems. We work with mid-sized manufacturers across Europe and team up with ERP and MES providers to bring this tech directly into real production workflows. We’re a lean, international team with big ambitions. Backed by top investors and working closely with global industry leaders, we’re on a mission to revolutionize manufacturing efficiency with cutting-edge AI tech.

• Design, develop, and deploy AI/ML solutions leveraging **LLMs, NLP, and Generative AI** • Build and optimize **Retrieval-Augmented Generation (RAG)** pipelines using vector databases and embedding models • Develop **agentic AI systems** (multi-step reasoning, tool use, orchestration frameworks) • Fine-tune, prompt-engineer, and evaluate large language models for production use cases • Build scalable, end-to-end ML/AI pipelines including data ingestion, preprocessing, model training, and deployment • Integrate AI solutions into applications via APIs and microservices • Collaborate with cross-functional teams (data engineers, product managers, and business stakeholders) to define AI-driven solutions • Implement **model monitoring, evaluation frameworks, and guardrails** (bias, hallucination mitigation, safety) • Optimize models and pipelines for performance, scalability, and cost-efficiency in cloud environments • Translate complex AI outputs into actionable insights for both technical and non-technical audiences • Contribute to AI best practices, architecture decisions, and internal tooling • Mentor junior engineers and guide teams on modern AI development patterns

Position SummaryThe Machine Learning Engineer will be responsible for the end-to-end development and deployment of Large language and machine learning models, with a primary focus on data preprocessing, model training, and fine-tuning using large-scale healthcare datasets. This role requires a strong understanding of Large language models, machine learning principles, data engineering, and experience working with sensitive healthcare data. Key Responsibilities - Data Preprocessing: Clean, transform, and prepare large, complex healthcare datasets for machine learning model development. This includes handling missing values, outlier detection, feature engineering, and data normalization. Identify, collect, and curate relevant, industry-specific datasets for model retraining. Format data appropriately for the chosen LLM and training pipeline - Model Training & Fine-Tuning: Design, train, and fine-tune various LLMs on extensive healthcare data to solve specific clinical or operational problems. Set up and manage the training environment, including GPU instances and required software. Train and fine-tune pre-trained LLMs on the custom dataset to achieve specific goals. Experiment with and fine-tune hyperparameters such as learning rate, batch size, and training epochs to optimize model performance. Integration of structured + unstructured data (multi-modal/multi-input models) - Model Evaluation & Optimization: Evaluate model performance using appropriate metrics, identify areas for improvement, and implement optimization strategies. - Pipeline Development: Develop and maintain robust and scalable data and ML pipelines for model training, inference, and deployment. - Collaboration: Work closely with data scientists, clinicians, and software engineers to understand requirements, integrate models into production systems, and ensure data privacy and security compliance. - Research & Development: Stay up-to-date with the latest advancements in machine learning and healthcare AI, and explore new technologies and methodologies to enhance our solutions. - Documentation: Maintain clear and comprehensive documentation of models, data pipelines, and experimental results.

• Build and operate production-grade model serving infrastructure using frameworks such as vLLM, TGI, Triton, or equivalent • Design and implement robust deployment pipelines with blue/green and canary rollout strategies for ML models • Develop and maintain auto-scaling systems, multi-model serving architectures, and intelligent request routing layers • Optimize GPU utilization, memory efficiency, network throughput, and model artifact storage performance • Design observability systems for tracking inference latency, throughput, GPU usage, cost metrics, and system health • Manage model registries and CI/CD pipelines enabling automated and reproducible model deployments • Own the full lifecycle of ML systems from development through production, including operational support and on-call responsibilities • Define engineering best practices and contribute to platform scalability in a fast-moving startup environment