Machine Learning Engineer, Covariant
3 days a week in office (Emeryville, CA)
AI Robotics company
Job no longer available
AI Robotics company
101-200 employees
Job no longer available
3 days a week in office (Emeryville, CA)
101-200 employees
Company mission
To build the Covariant Brain, a universal AI to give robots the ability to see, reason and act on the world around them.
Role
Who you are
- Bachelor’s degree or foreign degree equivalent in Computer Science or a related field, and five (5) years of progressive post-baccalaureate experience in Software or Robotics or in the job offered or related role
- Alternatively, the employer will accept a Master’s degree or foreign degree equivalent in Computer Science or related field and two (2) years of experience in Software or Robotics or in the job offered or related role
- Employer will accept any suitable combination of education, experience, or training
What the job involves
- Analyze failures and deploy improvements to models and robot logic
- Build tools to facilitate analysis and improvement
- Iterate all aspects of the Brain (models, algorithms, tooling, etc.) to resolve production failures and deliver human-level autonomy
- Build tools and analyze highly complex system of robots, models, components and sensors to understand errors and deploy performance improvements
- Analyzing robot performance for different errors, understanding their root causes, testing/deploying improvements to production, and building tools to help aid in the analysis
- Collaborate closely with research, Software, Hardware, and infrastructure teams to deliver highly reliable robotic applications
- Work closely with the Hardware team in station design optimization. Identify throughput bottlenecks in the application and work with the Software team to pinpoint and resolve them
- Take experimental models from the research team, test them to validate improvement, and deploy them to solve application problems
- Must appear in office 3 days per week; work from home permissible 2 days per week
- Software Engineering to understand codebase and make code changes
- Machine Learning to develop models that learn from data
- Statistics and Probability to run experiments and evaluate the performance of nondeterministic systems
- Deep Learning and Neural Networks, such as Rest, Graph Neural Networks, and Multilayer Perceptrons, building deep learning models and Convolutional Neural Networks to incorporate into company’s software
- Compilers and Programming Languages such as Pytorch, Tensorflow, and Numpy, to debug problems in graph-based frameworks
- Develop scoring algorithms to solve robotics problems and plan algorithms such as CMA-ES to control its motion; and
- Programming Systems such as distributed file system on Amazon S3 and programming hardware components such as conveyers, sorters, lifts, light curtains and sensors to debug problems related to file systems
Otta's take
Theo Margolius
COO of Otta
Covariant's investors have solid backgrounds in AI research and academia and have a clear vision of how they want to impact the robotics industry. They have successfully deployed of the Covariant Brain in different sectors, including fashion, health, logistics, and pharmaceuticals.
Robotic AI has made huge leaps in recent years, but a global labour crunch means that supply chain and manufacturing firms are demanding more from their robotic workforces, requiring a precision and dexterity not yet reached without human assistance. Covariant exists to provide just this, outfitting robots with a universal AI for robotic manipulation.
With the global robotics industry expected to balloon to $86 billion by 2027, there is no shortage of firms working in this field. Competitors include Kindred and RightHand Robotics. However, Covariant is unique in its focus on autonomous robotics and in the range of its applications, which can be used in formerly difficult areas such as order picking regardless of the type of objects that need to be manipulated. It also has an edge in that its solution can be deployed to existing robots without new hardware or extensive reprogramming, significantly reducing onboarding costs.
Covariant has received significant funding for its product and is expanding internationally and into new verticals. It is looking to further develop the Covariant Brain and to maintain its market lead as companies continue to double down on investments in automation.
Insights
Company
Funding (last 2 of 5 rounds)
Apr 2023
$75m
SERIES C
Jul 2021
$80m
SERIES C
Total funding: $222m
Company benefits
- Vision and dental insurance
- Flexible working hours
- Equal opportunity employer
- Work from home opportunities
- Health insurance
- Unlimited PTO
- Lunch and dinner each day (for on-site employees)
- Monthly health & wellness budget
- Quarterly learning budget
- 401k plan and match
Company values
- Breaking new ground, together
- Learning constantly
- Striving for empathy
Company HQ
Southwest Berkeley, Berkeley, CA
Founders
Tianhao Zhang
(Co-founder)Tianhao worked as a Grad Researcher during their UCB PhD, before co-founding Covariant in September 2017.
Pieter Abbeel
(President & Chief Scientist)Having taken a PhD at Stanford, Pieter started working as a Professor at UCB in 2008, before co-founding Gradescope and working as an OpenAI Research Scientist. They co-founded Covariant in 2017, where they have worked since.
Peter Xi Chen
(CEO)PhD in Philosophy and a BA in Computer Science from Stanford. Started his career as a Research Scientist at Open AI before co-founding Covariant.
Salary benchmarks
We don't have enough data yet to provide salary benchmarks for this role.
Submit your salary to help other candidates with crowdsourced salary estimates.
Share this job
View 8 more jobs at Covariant