About Lumaril

Lumaril is an industrial autonomy company that increases output at existing plants. Our platform connects to a facility's control systems, learns how the operation performs at its best, and continuously optimizes production for higher throughput, better yield, and more reliable operations.

We’re actively building a culture that fosters productive collaboration on difficult problems with real-world consequences. We’re early-stage, moving fast, and looking for people who thrive with high autonomy (and high ambiguity).

What You’ll Do

We’re looking for an engineer to build and deploy our systems for real-time control.

While Lumaril is first and foremost a software company, we integrate directly with our customers’ existing hardware, deploy off-the-shelf compute and sensing systems, and work with external vendors to design and integrate third-party hardware at customer facilities. These are systems already generating revenue. Changes to control directly impact output, yield, and efficiency.

We treat industrial facilities as partially observable, dynamic systems. By applying robotics techniques (perception, state estimation, control, simulation), we make these systems observable and controllable.

We begin with classical methods to deliver immediate, reliable improvements and expand across industrial control domains. As we integrate more systems, we capture high-quality data and build increasingly accurate models.

This foundation enables us to apply modern AI techniques to develop control policies that are more robust, adaptive, and generalizable over time. Ultimately, we believe control can be learned in a way that generalizes across hardware and domains, forming the basis for a true foundation model for industrial systems.

You will:

Work directly with customers to understand real-world system constraints, decompose problems into tractable subsystems, and build and deploy solutions in production environments
Collaborate with domain experts (sensor vendors, hardware partners, etc) to design and integrate complete systems
Build core systems for real-time control, including perception over visual and PLC data, state estimation, and control (model-based, feedback-driven, and learned policy)
Develop and maintain the infrastructure and libraries that power closed-loop decision-making in industrial environments
Deploy systems that directly adjust control logic and influence real-world output, yield, and efficiency
Balance rapid customer delivery with long-term product development, turning one-off solutions into reusable, scalable components

Who You Are

Requirements

2-5 years of experience applying robotics techniques such as perception, state estimation, controls, or simulation, gained in industry, research, or a combination of both
We welcome candidates from a range of backgrounds, including robotics, autonomous vehicles, defense and aerospace, industrial automation, and physical AI companies, as well as recent PhD graduates, master's graduates with relevant research or project experience, and postdoctoral researchers
Familiarity with robotics tooling, simulation environments, or industrial control systems (PLCs, SCADA)
Comfort debugging across the full stack, from sensors and hardware to software and data
Clear, articulate communicator who can explain complex technical concepts to engineers, operators, and executives alike
Works well with others, takes feedback seriously, and moves quickly from discussion to execution
Ability to travel approximately 15-35% of the time, including team colocations, periodic visits to our NYC and El Segundo offices, and customer site visits as needed

Nice to have

A graduate degree (MS or PhD) in robotics, mechanical or electrical engineering, physics, controls, or a related field grounded in dynamics, controls, and systems, or equivalent professional experience building robotics or physical AI systems at a robotics or software company
Experience with industrial or physical infrastructure, edge deployment, or applying ML to physical systems
Experience taking robotics or controls work from concept to deployment on real hardware or in real-world environments
Takes ownership of ambiguous problems and drives them to resolution in a fast-moving environment

Details

New York (Flatiron) or Los Angeles (El Segundo)
Team colocation every six weeks

Benefits

Daily lunch at the office

Fully paid medical, dental, and vision for employees and families

401k through Betterment