Blue Origin
Applied Machine Learning Scientist II (AI/ML)
Blue Origin, Seattle, Washington, United States, 98101
Enterprise Technology Ai Engineer
At Blue Origin, we envision millions of people living and working in space for the benefit of Earth. We're working to develop reusable, safe, and low-cost space vehicles and systems within a culture of safety, collaboration, and inclusion. Join our team of problem solvers as we add new chapters to the history of spaceflight! This role is part of Enterprise Technology (ET), where we're developing the digital infrastructure needed to build the road to space, with an emphasis on digital capabilities required to advance Blue Origin's mission. Join us to develop innovative AI models for our Rocket Factory with a collaborative team. We are looking for someone with deep technical expertise that is interested in positively impacting safe human spaceflight. If you are a passionate and high agency individual (have energy to independently drive initiatives, fix problems and improve status quo), we would love to hear from you. We encourage you to apply even if you do not believe you meet every single qualification. We are building an inclusive team and understand that learning curves exist for everyone. Role and Responsibilities: As part of the core AI team within the Enterprise Technology group, you will be instrumental in improving the manufacturing processes of space crafts. By helping build novel AI systems that harnesses multimodal information from disparate sources and automatically transforming them, your work will assist manufacturing engineers in our factories, and make them more efficient, reduce latency, and increase velocity of launches. Here are a few common tasks that are expected to be completed as part of your role: Design experiments, hypotheses and rigorously benchmark solutions Conduct literature review to identify relevant datasets, models and tools to prototype and improve AI systems in production Develop techniques for data preprocessing, extraction, and synthetic generation Deploy algorithms by creating docker images, spinning up containers, creating endpoints and microservices Document your findings and present to stakeholders Desired Background and Qualifications: Machine Learning, Masters or PhD with relevant course work - Linear algebra, Statistics, NLP, and Computer Vision, 3+ years of experience in the industry working on real-world data and building AI systems, 4+ years of experience coding - Python and AI frameworks (TensorFlow/ PyTorch), Background in Deep Learning - vision-language models (VLMs), transformers, training LoRA adapters, model distillation, and fine-tuning, Strong mathematical foundations - linear algebra, calculus, statistics, Ability to curate large-scale multimodal datasets - data preprocessing and generation of datasets (synthetic and human-labeled), Experience with ML libraries and tools
pandas, torchvision, NLTK, spaCy, diffusers, vllm, transformers, DSPy, Ray, LangGraph, DeepSpeed etc., Designing machine learning experiments
understanding and improving model behaviors and proposing solutions for generalization/interpretability, Familiarity with reinforcement learning principles, Understanding of AI ethics and responsible AI development practices, Distributed Systems, Expertise with GPU optimization and Cloud platforms (AWS, Azure), Experience with container orchestration tools (Docker, Kubernetes), CI/CD pipeline implementation and management, Systems architecture design for AI applications. Preferred Qualifications: Prompt and context engineering - experience with implementation of Retrieval-Augmented Generation (RAG), Proficiency in generative model architectures - GANs, VAEs, Diffusion Models, Experience with post-training techniques - instruction tuning, preference modeling, or RLHF, Experience with 3D modeling and building multi-agentic workflows, Demonstrate concrete achievements from ML projects that involve building multimodal algorithms for some of the following tasks: segmentation, object detection, classification, question and answering and so on, Interest or experience in test-time/inference scaling, Research publications in machine learning conferences, journals and/or open-source contributions.
At Blue Origin, we envision millions of people living and working in space for the benefit of Earth. We're working to develop reusable, safe, and low-cost space vehicles and systems within a culture of safety, collaboration, and inclusion. Join our team of problem solvers as we add new chapters to the history of spaceflight! This role is part of Enterprise Technology (ET), where we're developing the digital infrastructure needed to build the road to space, with an emphasis on digital capabilities required to advance Blue Origin's mission. Join us to develop innovative AI models for our Rocket Factory with a collaborative team. We are looking for someone with deep technical expertise that is interested in positively impacting safe human spaceflight. If you are a passionate and high agency individual (have energy to independently drive initiatives, fix problems and improve status quo), we would love to hear from you. We encourage you to apply even if you do not believe you meet every single qualification. We are building an inclusive team and understand that learning curves exist for everyone. Role and Responsibilities: As part of the core AI team within the Enterprise Technology group, you will be instrumental in improving the manufacturing processes of space crafts. By helping build novel AI systems that harnesses multimodal information from disparate sources and automatically transforming them, your work will assist manufacturing engineers in our factories, and make them more efficient, reduce latency, and increase velocity of launches. Here are a few common tasks that are expected to be completed as part of your role: Design experiments, hypotheses and rigorously benchmark solutions Conduct literature review to identify relevant datasets, models and tools to prototype and improve AI systems in production Develop techniques for data preprocessing, extraction, and synthetic generation Deploy algorithms by creating docker images, spinning up containers, creating endpoints and microservices Document your findings and present to stakeholders Desired Background and Qualifications: Machine Learning, Masters or PhD with relevant course work - Linear algebra, Statistics, NLP, and Computer Vision, 3+ years of experience in the industry working on real-world data and building AI systems, 4+ years of experience coding - Python and AI frameworks (TensorFlow/ PyTorch), Background in Deep Learning - vision-language models (VLMs), transformers, training LoRA adapters, model distillation, and fine-tuning, Strong mathematical foundations - linear algebra, calculus, statistics, Ability to curate large-scale multimodal datasets - data preprocessing and generation of datasets (synthetic and human-labeled), Experience with ML libraries and tools
pandas, torchvision, NLTK, spaCy, diffusers, vllm, transformers, DSPy, Ray, LangGraph, DeepSpeed etc., Designing machine learning experiments
understanding and improving model behaviors and proposing solutions for generalization/interpretability, Familiarity with reinforcement learning principles, Understanding of AI ethics and responsible AI development practices, Distributed Systems, Expertise with GPU optimization and Cloud platforms (AWS, Azure), Experience with container orchestration tools (Docker, Kubernetes), CI/CD pipeline implementation and management, Systems architecture design for AI applications. Preferred Qualifications: Prompt and context engineering - experience with implementation of Retrieval-Augmented Generation (RAG), Proficiency in generative model architectures - GANs, VAEs, Diffusion Models, Experience with post-training techniques - instruction tuning, preference modeling, or RLHF, Experience with 3D modeling and building multi-agentic workflows, Demonstrate concrete achievements from ML projects that involve building multimodal algorithms for some of the following tasks: segmentation, object detection, classification, question and answering and so on, Interest or experience in test-time/inference scaling, Research publications in machine learning conferences, journals and/or open-source contributions.