Computer Vision Engineer Lead

About The Role
As a Computer Vision Engineer Lead,  you will be working on building Machine Learning models to support our computer vision and data analysis efforts. The role will involve understanding our current systems and their limitations and bringing your expertise in augmenting and developing CV models. 
We are very interested in expertise around the subfields of semantic segmentation and instance segmentation and you will be involved in keeping up to date with the latest papers in these areas and implementing them using platforms like Tensorflow, or Pytorch. Robotics data also comes in a variety of different forms (time series, alerts, etc..) that are targets for models to help us predict and analyze behaviors emerging from our robotics systems. You will get to find and perfect the right model(s) and architecture, in addition to identifying key problems throughout our system that can be solved with computer vision techniques. 
About Dexterity: At Dexterity, we believe robots can positively transform the world.  Our breakthrough technology frees people to do the creative, inspiring, problem-solving jobs that humans do best by enabling robots to handle repetitive and physically difficult work.
We’re starting with warehouse automation, where the need for smarter, more resilient supply chains impacts millions of lives and businesses around the world.   Dexterity's full-stack robotics systems pick, move, pack, and collaborate with human-like skill, awareness, and learning capabilities.  Our systems are software-driven, hardware-agnostic, and have already picked over 15 million goods in production.  And did we mention we’re customer-obsessed?  Every decision, large and small, is driven by one question – how can we empower our customers with robots to do more than they thought was possible?
Dexterity is one of the fastest growing companies in robotics, backed by world-class investors such as Kleiner Perkins, Lightspeed Venture Partners, and Obvious Ventures.  We’re a diverse and multidisciplinary team with a culture built on passion, trust, and dedication.  Come join Dexterity and help make intelligent robots a reality!

Day to Day Activities:

• Experience working with RGBD datasets and point clouds as they are important to the type of problems you will encounter in the areas of computer vision for robotics. 
• Write optimized code in c++ and python and understand limitations of languages and frameworks. 
• Build your own models, training pipelines and serving infrastructure to run a production system. 
• The end to end process of gathering data, augmenting data, studying performance and optimizing quality of the model, and deploying and troubleshooting models on the field will be under your purview. 

Desired Skills:

• Engineering degree in Computer Science, or a related discipline, or equivalent experience.
• Strong knowledge of Modern C++ and Python.
• Experience creating Tensorflow and Pytorch graphs.
• Experience using profilers and debuggers to optimize code.
• Experience building and maintaining production code.
• Prior experience building and training production CV models from scratch using Tensorflow and pytorch.
• Experience with serving architectures like NVIDIA Triton and Tensorflow Serving, flask and others.
• Experience with cloud infrastructure such as AWS, GCP, Azure etc. and experience using this infrastructure to create training and serving pipelines.
• Experience deploying software and configuring distributed systems.