US Citizenship or Permanent Residency Required
Size Stream is a 3D body scanning company with applications in the apparel and health & fitness industries. We are seeking a talented DevOps Engineer to join our multi-disciplinary team. Our projects involve state-of-the-art technology and techniques, and we seek to push those boundaries even further.
Size Stream needs individuals who are excited about building a new, unique cloud-based platform that works with both 3D Body Scanners and mobile applications.
Work will involve improving our CI/CD systems, automating our monitoring and alerts, implementing changes to support a global reach, and automating where possible to keep the team agile and focused on the customer.
DevOps Engineer Responsibilities
- Apply cloud computing skills to automate deployment of new releases into various environments (Development, Staging, Production)
- Have a vision for the organization that will allow you to implement long-term solutions through simplification and automation.
- Troubleshoot production issues and coordinate with the development team to streamline code deployment
- Implement automation tools and frameworks (CI/CD pipelines)
- Implement alerting monitoring of production systems; including on-call schedules for teams alerts are intended for.
- Analyze code and communicate detailed reviews to development teams to ensure a marked improvement in applications and the timely completion of projects.
- Collaborate with team members to improve the companys engineering tools, systems and procedures, and data security.
- Conduct systems tests for security, performance, and availability.
- Provide high-performance cloud deployment solutions and maintain rapid deployment pipelines, server monitoring, and troubleshooting
DevOps Engineer Requirements
- You have 3+ years of experience as a DevOps Engineer or equivalent software-engineering role.
- You have experience working with cloud systems such as AWS, Azure, GKE (Experience with AWS is required)
- You have experience working with AWS Lambda functions.
- You are an expert in code deployment tools (Puppet, Ansible, and Chef).
- You have experience with building and deploying .NET applications.
- You have experience with CI/CD tools (Gitlab, Jenkins, Bamboo, etc)
- You have working knowledge of coding and scripting languages (Python, Ruby, Bash, Batch, PowerShell, Go)
- You have experience with artifact management tools such as Artifactory.
- You have a working knowledge of databases and SQL (Structured Query Language).
- Have knowledge of monitoring and logging systems such as Datadog, Splunk, Nagios, Zabbix, Prometheus, etc.
- Strong background managing Linux and Windows operating systems.
- You have strong knowledge of Git and Git workflows.
DevOps Engineer Preferred (Not required) Skills
- Experience with Kubernetes/Docker.
- Experience with Linux and Windows Containers.
- Experience using Conan, Make and CMake.
- Experience with PyPI
- Experience with C++ based build systems
We sat down with researchers at the University of Missouri to discuss how Size Stream 3D body scanners can help health and fitness professionals better track body composition and shape change throughout a person’s fitness journey.
It might not seem so strange to think, with all the crazy technology of the modern world, that we should have mastered something as simple as 3D scanning the human body. We have cars that can drive themselves, drones delivering pizza, and robots that can run and do backflips. Where are the picture-perfect body scanners giving world-class 3D detail?
To answer that question, you need to know what the challenges are. In other words, what is it that makes a scanner good and what makes a scanner so-so. If the purpose is to provide precise measurement for made-to-measure clothing, the scanner needs to be accurate enough that the end product is a perfect fit for the end user. Similarly, a body scanner used to monitor physical fitness needs to be able to track subtle changes in body shape over time. In principle, this means the scanner must be accurate to within a few millimeters (a millimeter is about the thickness of a dime).
Amongst 3D sensor applications, body-scanning is a strange beast. Some applications, like terrain mapping and collision avoidance for drones, can make do with errors of an inch or more. But if you buy a shirt with the collar 1 inch too short, how likely are you to buy that next shirt? Others, like manufacturing quality assurance and process control might require accuracy to 1/100th of a millimeter. Would you notice a sleeve 1/100th of a millimeter too long? Probably Not.
There are many potential solutions to the challenge of 3D body scanning. Different sensors and technologies provide the 3D data we need but come with tradeoffs that make picking the right tech as much art as science. In the extreme high-end, ultra-precise medical-grade 3D sensors give wonderful repeatable scans but come with a price-tag that puts them out of reach of all but the truly deep-pocketed. At the very low-end, sensors are cheap enough to be available to everyone but produce only modestly accurate scan data. So picking the right sensor is crucial to the successfully designing a scanner.
But it’s not just about sensor accuracy. Many 3D sensing applications require only a single sensor. But, with narrow fields of view, the small size of body scanners, and the large—in comparison—size of the human body, it takes many sensors to get the whole picture. Each sensor just contributes a part of the view.
Putting all this data back together is a little like being one of the king’s men in Humpty Dumpty. Each sensor has its own character and imperfections. Each sees things just a little bit differently. Each sees things at a slightly different point in time. The last one isn’t a problem so long as the scan subject remains perfectly still. No problem you say? Take a minute to record a video of yourself standing perfectly still. It’s not as easy as it sounds. A millimeter here, a millimeter there, and before you know it, the scan comes out looking like Frankenstein’s monster: a patchwork of poorly connected features.
Okay, now imagine you’ve solved the problem of putting all the bits together. Remember when we said each sensor has its own character and imperfections? It turns out that, in addition to being slightly distorted, each sensor varies in response to temperature, time, and lighting. That means that, from moment to moment, the data coming out of sensors jumps and drifts.
So, to summarize a bit: we’ve got a collection of sensors, each a little distorted, each seeing the scan subject from a different point of view, each changing in time, and we need to put them all together to match up perfectly. Quite a tall order, when you think about it.
Despite all this, Size Stream continues to produce the world’s most accurate body scanners for apparel, fitness, and uniform markets without breaking the bank. Our engineering team includes top talent from across the globe, with over a hundred years of combined experience. We apply the best minds in computer vision, numerical systems, artificial intelligence and machine learning, geometric body measurement, systems integration, cloud technology, and application design to deliver the utmost in 3D scan quality, innovative and engaging end-user experiences, maximum return on investment to our customers and partners.
Part of our continuing Customer Success series, we sat down to talk with Alton Lane about their experience with Size Stream’s 3D body scanner as a vital component of their men’s modern bespoke apparel store experience: