Job Summary:
X, The Moonshot Factory is focused on building innovative solutions that bridge foundational models and real-world industrial systems. As an AI Resident, you will collaborate with the core architecture team to design autonomous platforms that model physics and execute simulation code, pushing the limits of science and modeling.
Responsibilities:
โข Design Agentic Frameworks: Develop and scale multi-agent platforms that autonomously generate, test, and execute simulation code for complex physical, mechanical, or chemical systems.
โข Architect Knowledge Ontologies: Build the formal schemas, data structures, and physical ontologies that allow LLMs to reason accurately about machinery, physical constraints, and engineering primitives.
โข Close the Loop on Simulation: Integrate agentic code generation with active physics engines. You'll build the runtime verification and feedback loops that allow agents to self-correct when their simulation code fails.
โข Engage with the Community: Embedded directly within a confidential X project, you will collaborate with the broader X intern and AI community, participating in technical colloquia and talks with AI leaders.
Qualifications:
Required:
โข PhD Candidate in STEM: Currently enrolled in a PhD program (Computer Science, Physics, Engineering, Math/Stats) with a strong publication or research footprint in ML and excellent general-purpose programming skills.
โข Agentic Systems Experience: Hands-on experience building LLM systems that use multi-agent coordination, advanced tool execution, and complex context management. You should deeply understand how state-of-the-art coding agents (e.g., Claude Code, Gemini CLI, Codex) operate under the hood.
โข Ontologies & Knowledge Representation: A strong grasp of how to translate physical concepts, engineering rules, or hardware constraints into structured schemas and relational graphs that LLMs can reliably parse and act upon.
โข Simulation & Physical Modeling: Experience working with physical applications, hardware logic, or modeling real-world environments where code interfaces directly with physical variables.
Preferred:
โข Open-source contributions or peer-reviewed publications at top-tier venues (NeurIPS, ICML, ICLR, CVPR, etc.).
โข Experience building deterministic verification loops or execution environments where LLM agents learn from runtime errors.
โข Familiarity with physics engines, robotics simulators (e.g., MuJoCo, PyBullet, Gazebo), or CAD APIs to build digital twins and representations of physical machinery.
Company:
X, the moonshot factory invents and launches breakthrough technologies and radical solutions. It is a sub-organization of Alphabet. Founded in 2010, the company is headquartered in Mountain View, USA, with a team of 201-500 employees. The company is currently Growth Stage.