Data Scientist, ML Evaluation and Autonomy

Key Skills

AI technology, Autonomy, Cloud computing, Data science techniques, Machine learning techniques, Python Programming, SQL

Job Description

At Shield you’ll be part of an elite team that uses machine learning to develop the best AI pilot. Help us build world-class infrastructure for autonomous behavior, scene understanding and embedded AI. We develop large pipelines for training, evaluating and productizing machine learning models. As a member of our machine learning team you will work with colleagues to evaluate and improve new models in the areas of reinforcement learning, attention mechanisms and object detection and tracking. You will evaluate workflows to evaluate models and to ensure that they operate at peak performance onboard the different autonomous platform.

What you’ll do:

• Work closely with Subject Matter Experts and key stakeholders to define success, develop metrics and processes, and evaluate agent and team performance in a variety of scenarios
• Identify metrics to quantify agent performance in multi-agent adversarial scenarios of incomplete or imperfect information
• Own metrics reporting and distribution, in person and through automated reporting tools
• Test ML/RL model for regressive behavior
• Use rigorous statistical methods to generate synthetic simulated datasets for training the ML/RL models
• Contribute to the development of stable and unified evaluation code and pipelines that interact with existing company reporting tools

Projects that you might work on:

• rigorous evaluation of ML and RL models in simulated environments
• create efficient ways to represent and learn in multi-agent 3D environments
• maximize the efficiency of simulator-based training
• strive to improve end-to-end learning pipelines
• identify efficient approaches to transfer ML models to real-world platforms
• drive towards scale by leveraging low-fi/high-fi efficient training and other forms of agent pre-training

Salary Range: $115,440 $180,560 Annually + Bonus + Benefits+ Equity

Actual compensation is influenced by a wide array of factors including but not limited to skill set, level of experience, licenses and certifications, and specific work location.

Required qualifications:

• Bachelors in physics, engineering, mathematics, etc.
• You have a demonstrated record of working hard, being a trustworthy teammate, holding yourself and others to high standards, and being kind to others.
• Good written and verbal communication skills

Preferred qualifications:

• Masters degree in data science, machine learning, mathematics, physics or computer science.