Software Engineer, AI Product (Andrew K.)

Location: San Carlos, CA

Employment Type: Full time

Location Type: On site, 5 days a week

Department: Software Engineering, Behavior team

Compensation: [Senior and Staff bands not provided. Confirm with Carla], offers equity

About 1X

We're building humanoid robots that work in the home. Doing the chores, handling the tasks, and giving people their time back. Simple, but it's not.

To do this right, we have to solve robotics, AI, and manufacturing at the same time, at scale, in a form factor safe enough to live with your family. We've been at this since 2014. If you're inspired by this, you'll thrive here.

NEO is our flagship: a home robot designed to move, learn, and operate in the real world alongside real people. We're not demoing it. We're shipping it.

If you've spent your career on problems that matter and want to see them reach the world, this is that moment. We're scaling, we're hiring with intention, and we need people who want to build something that changes how humans spend their time.

About the Team

The Behavior team builds the software that makes NEO an intelligent, autonomous character living alongside people. We develop the decision making systems, world simulation, and behavior frameworks that let robots perceive their surroundings, understand context, prioritize goals, and interact naturally in dynamic environments.

NEO talks with you the way an LLM would. Behind that conversation sit several models: a conversational LLM, a world model that understands the space around the robot and takes high level commands like "walk to the door, open it, and go outside," and the motion systems that carry them out. The Behavior team builds the layer that turns what a person says into commands those models can execute, and keeps the whole system on task.

This work runs partly on the robot and partly in the cloud. You work closely with the Product Software team building the cloud platform, and on site next to the hardware, embedded, and AI teams building our world model. CI runs with hardware in the loop, so what you build gets tested on real robots.

Your Charter

Build the behavior systems that bring humanoid robots to life. You will design scalable AI architectures that let robots reason about their environment, make autonomous decisions, manage competing goals, and perform complex tasks in real homes.

At the core is an orchestration framework for the robot. It takes user input, decides what the conversational LLM should do, manages its memory and context, and provides executive function that keeps it on track. It breaks a request into steps the world model and motion systems can execute. The closest analogy is an LLM agent orchestration framework, except the agent has a body and acts in the physical world.

You write code on the robot and in the cloud, and you own how those pieces fit together. We build with AI coding agents every day, so the bar is system design and ownership: design the architecture, direct the agents, review what they produce, and stand behind what ships.

Key Outcomes

• Build an orchestration framework that interfaces with multiple LLMs and custom models, turns natural language requests into high level robot commands, and breaks complex tasks into simple executable steps. This is the foundation for robots that perform household tasks naturally and autonomously
• Build the memory and context layer: systems that extract, store, and index relevant details about environments and the people the robot lives with, and feed the right context to each model at the right time
• Build executive function and decision making frameworks using behavior trees, hierarchical state machines, planners, or similar architectures, so the robot manages competing goals and stays on task
• Split behavior logic cleanly between on robot and cloud components, and integrate with the Product Software team's cloud platform
• Partner with Core, World Model, Apps, and Motion teams to put behavior systems into real world robot operation
• Improve the realism, robustness, and adaptability of robot behaviors, and set the engineering practices for building behavior at scale

Key Competencies

• Shipped AI products. Has built products on top of LLMs and custom ML models, and exposed those capabilities in intuitive user experiences. Has used LLMs to carry out real multi step workflows, not just chat
• Agent systems fluency. Hands on with conversational interfaces, agent orchestration, prompt engineering, MCP, tool use, and management of agent context and memory
• Decision making architecture. Strong understanding of planning systems, behavior trees, state machines, and autonomous agent behavior
• Systems thinker. Simplifies complex autonomous behavior into maintainable software architecture, and designs across the robot and cloud boundary
• Excellent engineer. Strong software engineering fundamentals. Uses AI coding agents to move faster and understands every line that ships
• Technical leader. Comfortable leading technical initiatives and working across ML, robotics, and software teams in a fast moving environment

Minimum Requirements

• 6+ years (Senior) or 10+ years (Staff) of professional experience building conversational AI, NPC or game AI behaviors, or system architecture
• Shipped at least one product that integrates LLMs or ML models into a real user workflow
• Strong systems programming and software design skills
• Bachelor's degree in Computer Science, Software Engineering, Robotics, AI, or a related technical field
• Able to work on site in San Carlos, CA

Preferred Skills

• Built agent orchestration, memory, or context management systems for LLMs
• Implemented behavior trees, GOAP, hierarchical state machines, planners, or similar AI architectures under tight performance constraints
• Experience building open world games, large scale simulations, conversational AI, or emergent gameplay systems
• Robotics, autonomous systems, reinforcement learning, or embodied AI experience
• C++ for on device code, plus cloud services experience
• Navigation, pathfinding, task scheduling, perception, or world modeling
• Built tools that let designers, researchers, or engineers rapidly author and debug complex AI behaviors