Senior Machine Learning Engineer, Parallel Domain
Pose Estimation
$150-180k
Synthetic data for computer vision and perception
Open for applications
Synthetic data for computer vision and perception
21-100 employees
Open for applications
$150-180k
21-100 employees
Company mission
To speed up the transition to autonomy by leveraging synthetic data.
Role
Who you are
- Advanced degree: MS or PhD in ML, computer vision, robotics, or related field
- Relevant experience: Industry experience applying ML to computer vision problems, with emphasis on pose or multiview geometry
- Technical proficiency: Strong Python experience and production-level code development
- Foundational knowledge: Deep understanding of 3D geometry, linear algebra, and optimization
- If you're passionate about machine learning, 3D reconstruction, generative AI, and the future of autonomous systems, we'd love to hear from you. Apply today and help us revolutionize the world of AI!
Desirable
- Hands-on experience: Demonstrated experience in developing and deploying production-level machine learning models
- Parallel Compute Experience: Experience in programming CUDA kernels for 3D computer vision applications
- Research background: Familiarity with academic research in 3D reconstruction, visual geometry transformer models, bundle adjustment or related fields
- Industry knowledge: Understanding of the autonomous systems landscape and the potential applications of machine learning in this domain
- Publication record: Publications in top-tier conferences or journals related to machine learning
What the job involves
- We are looking for a Senior Machine Learning Engineer focused on pose estimation and structure-from-motion (SfM) to improve the backbone of Replica’s geometric reconstruction pipelines
- You will develop and integrate ML-based and classical approaches for robust camera pose estimation and scalable multiview geometry systems
- Build robust SfM and pose pipelines: Improve accuracy, speed, and reliability of pose estimation components
- Develop geometric vision models: Design and implement learned modules for camera calibration and pose refinement
- Integrate ML and classical methods: Combine geometric algorithms with modern ML tools for practical performance
- Collaborate on data and evaluation: Partner with ML engineers to curate datasets and define metrics for success
Our take
Autonomous vehicles are fast making the leap from science fiction to reality. However, there are huge safety implications that need to be addressed. Parallel Domain has developed software for training autonomous vehicles in a virtual setting, reducing the risks associated with real-world testing and giving vehicle manufacturers the ability to identify problems and put them right, before the vehicle ever hits the road.
Founded in 2017, Parallel Domain leverages synthetic data to train and test the vision and perception systems for AVs. Clients can customise training environments with parameters such as weather, lighting, and distribution of agents, allowing them to model specific situations, including rare ones that would be unlikely to occur in real-world training. This results in AIs that behave better and more safely in novel situations.
Parallel Domain has seen significant investment that will fuel operational expansion and increase business reach. With the value of synthetic data quickly becoming apparent, the company is positioned well to benefit from those wishing to train their autonomous vehicle AIs faster and more efficiently.
Steph
Company Specialist at Welcome to the Jungle
Insights
62% employee growth in 12 months
Company
Funding (last 2 of 3 rounds)
Nov 2022
$30m
SERIES B
Dec 2020
$11m
SERIES A
Total funding: $43.9m
Company benefits
- Flexibility to work from our office in the San Francisco Bay Area or your home office
- Competitive compensation
- Employer-paid supplemental medical, mental health, dental, and vision benefits
- 401(k)
- Paid vacation and sick time, winter shutdown, and 11 stat holidays each year
- Paid parental leave
- New hire equipment + accessories budget to optimize your setup
- $1,500 annual learning and development allowance
Company HQ
Financial District, San Francisco, CA
Leadership
Kevin McNamara
(CEO)They studied Computer Science at Harvard before working for almost 3 years as a Microsoft Technical Artist. They left their Procedural Content Lead at Apple in July 2017 to found Parallel Domain, where they have served as CEO since.
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 2 more jobs at Parallel Domain