Edge Performance Software Engineer

San Francisco, CA, United States


Role Location

  • San Francisco, CA, United States


11 - 25 people


320 Alabama St Apt 2
San Francisco, CA, 94110, US

Tech Stack

  • Go
  • Python
  • Node.js
  • Docker
  • React
  • Angular.js
  • AWS
  • CUDA
  • C++
  • Tensorflow

Role Description

Aquabyte is seeking an Edge Performance Software Engineer in San Francisco to work on improving the performance of our algorithms running on our hardware deployed on-site in Norwegian fish farms. You will be responsible for increasing our total analytics output per farm by increasing frames captured per second, filtering non-useful data, compressing data, optimizing the research team’s algorithms, and processing the most priority data when resources are exhausted. We deploy a chain of processing stages starting at the underwater enclosure, the fish pen the enclosure is in, the central floating barge the pens are attached to, and finally in the cloud.

You should be familiar with Linux, image processing, networking, and developing software for both ARM64 and AMD64 architectures. Bonus points for experience utilizing GPUs for both image processing and machine learning inference.

Aquabyte helps fish farmers manage their day-to-day operations by using off-the-shelf cameras run through computer vision and machine learning algorithms to provide biomass estimation, sea lice counting, and optimal feeding schedules.

Job Responsibilities Increase the analytic throughput of Aquabyte algorithms from image capture to dashboard result. Deploy software to remote Internet-connected farms safely. Develop performance testing frameworks to ensure that when algorithms are altered or new algorithms are developed that we have enough hardware resources available to produce a statistically significant amount of data per pen running these algorithms. Help the machine learning team produce algorithms and models that run faster. Work with optics team to help improve the sharpness of our imagery through dynamically changing camera settings adapting to real-world conditions. Dig into c libraries, python libraries, golang libraries that we utilize to squeeze out extra performance for our target deployment architectures.

About Aquabyte

We are one of the few companies applying machine learning / computer vision to directly solve the world’s food sustainability issues.

Fish farming is the #1 fastest growing sector of food production - a $160B worldwide industry. By improving fish farm efficiency, we help close the world’s impending protein deficit.

We work on very challenging distributed image processing and machine learning problems, and our work has direct relevance on the world’s food chain. Come join us!

Apply now: https://jobs.lever.co/aquabyte

Company Culture

We rely on our intuition and intention - We reward intellectual curiosity and debate. We encourage people to work on projects they find worthwhile as they may yield products and ideas that may add enormous value. We encourage people to be aware of what they are building rather than merely going through the motions - People should be learning in their current roles and learn about other roles in the organization. This helps share and build intuition across the company. A minimum of 5 - 10% of our time should be spent learning. - We believe that the best products are built from first principles and intuition rather than from someone telling us what to do. This is a harder road to take, but it will allow us to build a better product overall and allows us to scale the team laterally.

We are pragmatic and maximize efficiency - We value getting something done in a practical manner. Often this means doing 20% to get 80% of the value, and prioritizing the 20% over time. - We aim to maximize efficiency of everyone’s time. Often, this means having work-life balance such that we can work optimally and not get burned out. Take time off when you want, work when you want, but be optimally efficient. - We share information freely and everyone has access to everyone in the company, there are no secrets. By everyone knowing, everyone is able to develop intuition and work better together - We should over-communicate to ensure that things are done efficiently, and issues are triaged. We should speak and engage and disagree tactfully.

We believe in individual exceptionalism - Everyone on our team should be entrepreneurial in their own right and treat the company as their own. Projects should be championed and promoted laterally across the organization instead of top-down. - Everyone on our team is expected to push the envelope on what they are working on. If we hire exceptional folks everyone is held to a high standard. - Everyone is nice and kind to each other.

Interested in this role?
Skip straight to final-round interviews by applying through Triplebyte.