Staff Software Engineer - Training and Metrics

Meet the team:  

In the Training and Metrics group our core obligation is to develop and own the production-grade automated processes of the perception data loop. Starting with data selection, following training, evaluation and deployment, in the end, our machine learning models make our trucks perceive the world! We enable automated, data driven iteration cycles on the perception machine learning stack. To be successful, we collaborate with different perception teams, as well as Data Science and ML Ops teams across the company. 

What you'll do: 

• Lead the development and maintenance of the automated perception training infrastructure (AWS) and software stack (PyTorch) 
• Integrate and connect data selection (active learning), machine learning training and evaluation for perception multitask and fusion models 
• Accelerate the optimization of the perception machine learning software stack (PyTorch) with regards to training time, network usage and GPU/CPU utilization 
• Lead the unification process of common platform, developer and pipeline tooling such as data loading, modular network architectures, model conversion for embedded hardware and evaluation processes and common data formats 
• Coordinate and negotiate between multiple perception teams 

What you'll need to succeed: 

• Master's degree in computer science, data science, artificial intelligence or related field 
• 3+ years of software engineering experience using AWS 
• 5+ years of machine learning engineering focus in the field of perception 
• Knowledge and application experience of deploying SE best practices like coding standards, formatting and test-driven development 
• 7+ years of professional experience 
• 2+ years of experience with python and PyTorch, ideally on AWS 
• Experience in working within a data loop and deploying continuous trainings and data selection strategies 
• Coordination and communication skills in technical matters 

Bonus points: 

• PhD in a related field  
• Experience with:  
  • Kubernetes, Slurm
  • Docker / Jenkins
  • AWS Sagemaker, Batch
  • Unit testing of neural network (components) 

Perks of Being a Full-time Torc’r 
Torc cares about our team members and we strive to provide benefits and resources to support their health, work/life balance, and future. Our culture is collaborative, energetic, and team focused. Torc offers: 

• A competitive compensation package that includes a bonus component and stock options 
• 100% paid medical, dental, and vision premiums for full-time employees  
• 401K plan with a 6% employer match 
• Flexibility in schedule and generous paid vacation (available immediately after start date) 
• Company-wide holiday office closures 
• AD+D and Life Insurance