Physics-Based Modeling Engineer - Industrial AI

Company Description

TwinEdge is an early-stage startup building industrial AI that works without the cloud. We're bridging edge computing with cloud analytics to deliver digital twin capabilities for critical infrastructure, water treatment plants, HVAC systems, manufacturing facilities, and industrial operations. We combine first-principles physics with machine learning to give plant engineers real-time efficiency calculations, predictive maintenance, and anomaly detection that actually makes sense to the people operating the equipment.

We support 10+ industrial protocols (Modbus, OPC UA, Siemens S7, EtherNet/IP, MQTT, BACnet, DNP3, PROFINET, and more), run ML inference directly on edge hardware, and provide a full SaaS analytics platform, all built on the principle that industrial software should be practical, deployable, and understandable.

Role Description

We are looking for someone who thinks in terms of how machines actually work. Someone who looks at an analytical chart and sees the physics behind it. Someone who gets excited about connecting a sensor to a real piece of equipment and watching data flow into a model that predicts when something is about to go wrong.

This is a fully remote position, you will work with physical machines, real sensors, and actual industrial equipment. You will need to be comfortable getting your hands on hardware, understanding how things move, flow, spin, and break, and then translating that understanding into models and code.

What you will do

• Build and refine physics-based models for industrial equipment, motors, compressors, fans, conveyors, HVAC systems, rotating machinery, using first-principles engineering (conservation of mass, energy, thermodynamics, vibration analysis, heat transfer, mechanical dynamics)
• Work with real sensors and industrial hardware to validate models against actual operating data from machines in the field
• Develop anomaly detection logic that distinguishes real problems from sensor noise, using physics as the baseline and ML as the enhancement layer

Searching for Development & Programming roles that provide visa sponsorship? Connect with international employers through Development & Programming Jobs with Visa Sponsorship opportunities actively seeking talented professionals.

• Create digital twin simulations that let operators ask "what if" questions - what happens if a motor overheats, if load increases beyond spec, if we change the control strategy, if a bearing is degrading
• Translate complex physics into practical, interpretable analytics that plant engineers and operators can actually use
• Collaborate on building out our progressive ML pipeline, starting with physics-only models that work on day one, then layering in machine learning as operational data accumulates
• Work with industrial protocols and edge devices to ensure sensor data is accurate, reliable, and properly contextualized

What we are looking for

You understand physics and machines. Whether you learned it in school, on the job, or by taking things apart in your garage, you have an intuitive grasp of how physical systems behave. Fluid dynamics, thermodynamics, mechanical systems, electrical systems, you don't need to be an expert in all of them, but you need to be genuinely curious about how things work and why they break.

You can code. Language doesn't matter, what matters is that you can pick the right tool for the right problem. Need to crunch sensor data? Maybe that's Python. Need something performant on an edge device? Maybe that's Rust or C. Need to prototype a model fast? Maybe that's something else. We don't care what language you reach for, we care that producing meaningful results. You use the right tool to solve the problem at hand.

You're willing to learn fast. We don't care about your degree. We care about your ability to pick up new concepts, apply them, and iterate. If you don't know something, you figure it out. You read the manual, ask chatGPT or your favorite AI tool. You search for the paper. You ask the right questions. You build a prototype and test it.

You're comfortable with hardware. This role involves working with physical sensors, industrial equipment, and edge computing devices. You should be someone who's not intimidated by wiring a sensor, reading a datasheet, or troubleshooting why a Modbus register is returning garbage.

You think in systems. A pump doesn't exist in isolation, it's part of a system with pipes, valves, controls, and operating conditions. You naturally think about how components interact and how changes propagate through a system.

Location & Work Setup

• Based in: Portland, Oregon

Explore our comprehensive directory of visa sponsorship jobs from employers worldwide who are ready to sponsor talented international professionals.

• Work arrangement: Fully remote
• Important: While the position is remote, you will periodically work with physical machines, sensors, and industrial hardware. This may involve hands-on work with test equipment, sensor installations, or visits to industrial sites. You should be comfortable with this hybrid digital-physical work style.

Interview Process

We don't do whiteboard algorithms. We don't do trick questions. We give you a real problem and let you solve it the way you'd actually solve it on the job.

Here's how it works:

1. You'll receive a problem. It will be a practical, physics-and-engineering challenge, the kind of thing you'd encounter working on our platform. Think: given this sensor data and these equipment specs, figure out what's happening with this machine and build something that solves it.
2. Use any tool you want. Seriously, any tool. AI assistants (ChatGPT, Claude, Copilot, Grok whatever you prefer), search engines, documentation, Stack Overflow, textbooks, Python notebooks, MATLAB, Excel, pen and paper. The tools are not the test.
3. Solve the problem and explain your reasoning. We want to see how you think. How do you break down the problem? What assumptions do you make? How do you validate your approach? Can you explain the physics behind your solution? Can you identify where your solution might be wrong?
4. The goal is not perfection, it's competence and curiosity. We want to see that you can take an unfamiliar problem, use the resources available to you, arrive at a reasonable solution, and articulate why it works. That's the job.

Compensation

Compensation includes equity. As an early-stage startup, we offer meaningful ownership, Cash compensation is above average and depends on experience, we will be straightforward about numbers early in the conversation so nobody wastes their time.