Generative Spacecraft Design Software Engineer (ML/AI)

Job description

[https://app.ashbyhq.com/api/images/user-content/bdefb51d-7670-4c8c-9bae-c12592fbf204/cc5ceb76-e72c-48bb-a022-b5a9d56c5891/Wordmark_Oligo_v2.png] Oligo is building a manufacturing-in-the-loop foundation model to automate spacecraft design and production worldwide. Our approach allows customers to focus entirely on their own technology and mission objectives, while we handle everything, from design and manufacturing to launch and operations. Leveraging cutting edge AI-driven generative design and automated manufacturing, our ex-MIT, Harvard, and NASA JPL team work to create the most advanced payload-specific spacecraft at scale in weeks over months. With world‑class advisors on our board, and fresh funding from top investors like Lux Capital, we’re always on the lookout for exceptional builders, fast learners, and ambitious engineers. Whether your passion lies in spacecraft systems, avionics, ML/AI, or advanced manufacturing, you’ll be collaborating across disciplines on real missions that fly, perform in orbit, and scale internationally. We pair world-class AI/ML talent with top-tier satellite engineers under one roof to reimagine how space systems are built, starting from first principles. No bureaucracy. No legacy thinking. If you think you’re a fit, we are extremely excited to meet you. ROLE OVERVIEW Oligo builds vertically integrated infrastructure for automated spacecraft design and manufacturing. Our AI software stack, Zenith, turns raw mission requirements into flight-ready spacecraft using a pipeline of agentic AI systems, embedded simulation, and hardware-in-the-loop validation. We’re hiring a Software Engineer with a background in ML/AI to help advance core algorithms used for generative design, simulation-aware geometry creation, and multi-domain system optimization. You’ll work alongside spacecraft engineers and flight software developers to build AI that doesn't just simulate reality—but designs it. This is a hands-on, high-leverage role for early-career engineers interested in applying cutting-edge AI to real-world hardware. You’ll learn spacecraft engineering, astrodynamics, and manufacturability by building models that directly influence how our satellites fly. WHAT YOU'LL DO - Develop and deploy vision-language models (VLMs) that parse technical documents, datasheets, and system specs into structured engineering constraints. - Build and train reinforcement learning agents that explore multivariable design spaces—balancing structural, thermal, orbital, and manufacturability objectives. - Design and run CNN-based topology optimizers (e.g., U-Net + FEniCSx) for reducing mass in structural components while maintaining stiffness and compliance. - Construct multi-agent AI systems that simulate, evaluate, decompose, and iteratively redesign spacecraft configurations using embedded physics models. - Contribute to our infrastructure for automated, high-context model training using flight data, past simulations, and test results—enabling real-time design intelligence. - Work on automated system decomposition engines that convert high-level mission goals into detailed subsystem design specifications. - Interface with physical simulation tools (Ansys, GMAT, Thermal Desktop) and CAD environments (OpenCascade, CadQuery) to ground designs in physical constraints. - Collaborate daily with engineers building the real hardware—what you code will be tested in thermal chambers, vibration tables, and flown on orbit. WHAT YOU'LL BRING MINIMUM QUALIFICATIONS - Bachelor’s/Masters degree (or final semester) in Computer Science, ML/AI, Engineering, or a related technical field (or willing to leave existing program) - Experience in ML/AI through research, personal projects, or internships (not just coursework). - 2+ years of experience building advanced ML systems (deep learning, RL, planning, or LLM agent frameworks). - Strong proficiency in Python, including PyTorch, JAX, or TensorFlow, and modern tooling (e.g., LangChain, Ray, FastAPI, DVC, Docker). - Experience working on one or more of the following: - Vision-language models (e.g., transformers, CLIP, Flamingo-like systems) - Reinforcement learning for constrained optimization or control - CNN architectures for image-based optimization tasks (e.g., SIMP replacement) - Simulation-aware ML pipelines, physics-informed ML, or surrogate modeling - Ability to think clearly about tradeoffs between simulation fidelity, inference speed, and manufacturability. PREFERRED SKILLS AND EXPERIENCE - Experience integrating with CAD kernels (OpenCascade, Onshape API, or SolidWorks API), FEA tools, or mechanical simulation frameworks. - Familiarity with prompt engineering, tool calling, memory routing, and multi-agent system design. - Hands-on ability to prototype, build, and debug hardware systems—bonus if you’ve worked with microcontrollers, sensors, or test rigs. - Willingness to work extended hours or weekends when necessary to meet mission-critical deadlines. - You’ve worked on engineering project teams, design/build competitions, or research groups focused on real hardware or systems. - Familiarity with spacecraft concepts, astrodynamics, or control systems is helpful but not required PAY RANGE - Salary range: $110,000 - $164,000 / per year. - This role is on-site in Hawthorne, CA BENEFITS - Equity - Unlimited PTO - Medical (Platinum coverage), Vision, & Dental Insurance - Catering provided on-site everyday. ADDITIONAL INFORMATION You may be eligible for our suite of benefits including medical, vision & dental coverage.