Member of Technical Staff, Machine Learning

present of AI; New York (NY); On-site; USD 140,000 - 180,000; Full time

Job summary

Antioch provides high-fidelity, cloud-scale physical simulation software to develop and test autonomous robotics systems. The role focuses on hands-on machine learning to align simulations with real hardware, emphasizing applied ML, statistics, and modeling physical systems such as sensors or actuators.

Proficiency in Python data pipelines and evaluation is expected. As part of a fast-moving team, you will assess sim-to-real gaps, calibrate digital twins, and build feedback loops where

Qualifications

• Hands-on ML position with strong applied ML and statistics skills.
• Background modeling physical systems such as sensors or actuators.
• Good Python habits for data pipelines and evaluation.

Job description

Member of Technical Staff, Machine Learning

The role is a hands-on machine learning position focused on making robot simulations match what happens on real hardware. A strong candidate has solid applied ML and statistics skills, some background modeling physical systems such as sensors or actuators, and good Python habits for data pipelines and evaluation. Comfort with messy real-world data and with shipping work in an early, fast-moving team matters as much as research depth.

How it fits the AI era

AI is moving simulation from hand-built physics rules toward models learned from data, including world models and diffusion models, and this role sits right at that shift. Coding agents are expected to be part of daily work, with careful checking of what they produce. Judgment stays human: deciding whether a simulated gain will hold up on real hardware, spotting data leakage, and weighing uncertainty for customers.

How it builds AI experience

Someone in this role would gain hands-on experience with learned simulation, calibration of digital twins, and building feedback loops where each customer deployment improves the models. To prepare, a candidate could calibrate a simple robot simulation in MuJoCo or Isaac Sim against real or logged data, and measure how big the gap is. Trying a small learned dynamics model next to a physics model is a good way to show the skill.

Worth showing

• Sim-to-real gap measurement
• System identification and calibration
• Applied ML shipped to production
• Python data and evaluation pipelines
• Careful validation of agent-written code

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Antioch

Antioch provides high-fidelity, cloud-scale physical simulation software to develop and test autonomous robotics systems.