image description

Life at Dextro

image description

Why we do what we do:

We build machine learning systems that make sense of the deluge of sparsely-tagged video created by users and cameras on the Internet every day. We excel on the hardest problem, which is shaky, blurry, real-world user generated video. Simply put, working on ML is fun, and seeing our research and engineering efforts powering major systems used by millions is incredibly exhilarating. We're excited by all the future applications of our work in robotics and IoT.

image description

Own your code.

We're a small team building a complex system. Every engineer owns a major part of our processing pipeline and is responsible for its architecture and design. We don't have legacy code, and new projects blend research and engineering. We thrive on discussing hard problems in computer vision, NLP, and data science, and love hacking together.

image description

What our teammates say:

“Every day, I come into our office knowing that I am part of an awesome team. Our technology will change the future of video search, and we revel in that. As a small team, each effort matters. There’s nothing like pushing through a long day and recognizing that your work has made an appreciable impact on your company’s success.” —Marj Berman, Operations Manager

Current Openings

  • Dextro's mission is to power the intelligence behind every video collection, livestream feed and connected camera. We believe in showing, not telling, people when explaining to them all the things you can do with Dextro. As our first Product Marketing Manager, you’ll sit at the intersection of the product, engineering, and sales where you’ll help grow our business to the next level.

    In this position you’ll be required to think creatively about how to market our capabilities to both a technical and business centric audience. You’ll have full ownership of our marketing strategy and will collaborate with our designers, engineering and PR teams to bring new ideas to life.

    Responsibilities
    • Develop and position Dextro’s voice and brand within the market such that it resonates with potential customers and helps drive growth
    • Own our marketing strategy
    • Write & develop copy for our websites, blog posts, and all sales collateral
    • Collaborate with our designers on our product designs
    • Conceptualize high impact proof-of-concepts and work with our engineering teams to make them come to life
    • Liaise with our marketing partners
    Qualifications
    • 4-5 years of experience in B2B product marketing
    • An eye for great design and experience with Adobe Suite
    • A technical background, including frontend stack coding (HTML and CSS)
    • Familiarity with APIs
    • Ability to break down difficult problems into simple concepts for a targeted audience
    • Able to approach product and messaging from both an intuitive and data-centric way
    • Excellent verbal and written communication skills with a sense of humor
    Bonus points
    • Experience marketing SaaS products
    • Experience with Ruby on Rails
    • Have worked at (or founded) a startup!
  • Dextro's video understanding system is incredibly compute-intensive — especially when processing live video. This necessitates a complex collection of low-latency machine learning services on the backend. As our first dedicated Systems Engineer, you will take us from great code to great systems. Systems that are scalable, fully automated, and, most importantly, reliable.

    You will hold a senior position (3-6 years of prior engineering experience) on our small but rapidly-growing backend team. You will have complete oversight of our compute systems. From scaling our systems to prepare for a new product launch to sharing feedback that you learn while diagnosis bugs, you will own the management of Dextro’s systems. You will maintain system uptime and alter architecture to improve our stability. In the process, your work will form the base of successful customer relations. If you are currently an SRE working in a small ecosystem, but looking for more impact and ownership, this is absolutely a position for you!

    Our backend is built in Python, CUDA, and C++; we make use of all relevant services in the AWS ecosystem.

    Responsibilities
    • Scale systems to handle thousands — yes, really thousands! — of hours of video at once
    • Build testing and deployment environments
    • Identify bugs and incorporate performance feedback into systems
    • Research and select tools to best power our systems
    Qualifications
    • Thorough understanding of infrastructure and scalability
    • Fluency with AWS stack
    • Experience designing intuitive, developer-friendly APIs
    • Ability to take complete ownership of projects, from architecting to implementation to testing and release
    • Strong opinions on software development best practices and what makes for good code.
    • No computer vision experience needed
    Bonus points
    • C/C++/CUDA a plus but absolutely not required.
    • GPU optimization experience.
    • Have worked at (or founded) a startup!
  • Dextro is a video analysis company that is trying to understand everything that is in a video. To achieve this we need to organize all the concepts, whether visually present or said. That’s hundreds of thousands of concepts to classify, and that number will continue to grow. Currently we process thousands of hours of video every day, both prerecorded and livestreamed. Our mission is to power the intelligence behind every video collection, livestream feed and connected camera. As a distributed systems engineer, you will help us turn that goal into reality.

    As a member of our small, but rapidly growing team, you'll build systems assets that impact our core product. Your work will touch the Dextro API and workers that allow us to digest and analyze immense quantities of visual data. We give our developers unparalleled agency and creativity to build code. Your playground will be anything between the website front page and vision models, and you will have complete freedom to build those systems as you see fit. You will solve challenges that enable our product to be smarter and faster — and thus enable us to reach our company goals. If you are currently an backend dev working in a small ecosystem, but looking for more impact and ownership, this is absolutely a position for you!

    Responsibilities
    • S3 optimization
    • Create system to storage and analysis usage and other metrics
    • Build out our API server and other systems
    • Take an idea from small scale concept to large scale real world applications
    Qualifications
    • Have built distribution systems
    • Experience designing intuitive, developer-friendly APIs
    • Fluency with AWS stack
    • Ability to helm development of others’ ideas, from architecture to testing and implementation
    • Strong opinions on software development best practices and what makes for good code
    • No computer vision experience needed
    Bonus points
    • C/C++/CUDA a plus but absolutely not required
    • Redis, Celery, Flask is icing!
    • GPU optimization experience
    • Previous industry coding experience
    • Have worked at (or founded) a startup!
  • At Dextro, our goal is to finally deliver on the promise of the semantic side of the computer vision field. As researchers ourselves, we started Dextro because we realized that only by building a commercial vision product with wide adoption can we reach the scale and data throughput necessary to achieve this goal. We've developed a great system so far and particularly made strides on video-specific approaches; let's work together to push the envelope further.

    Use our large, well-curated proprietary dataset and our chunky GPU cluster to develop your ideas and refine your approaches in the pursuit of making the Dextro system more accurate, comprehensive, and scalable. This position has a great deal of autonomy, but our focus is on shipping; unlike in a lab, all of the code you write and successful concepts you develop will go live immediately.

    If you are currently in a graduate program, we offer summer positions and semester positions as well.

    Responsibilities
    • Guide Dextro’s research in deep learning
    • Build and improve upon state of the art deep learning systems
    • Fuse visual and audio information into a complete understanding of video content
    • Take an idea from small scale research to large scale real world applications
    Qualifications
    • Extensive computer vision and deep learning research experience, specifically applied to video
    • Understanding of various deep learning tools and techniques like Convolutional Neural Networks, Recurrent Neural Networks, Long Short Term Memories
    • Experience in GP-GPU programming environments, eg: Cuda or OpenCL
    • Desire to apply your skills and experience to production systems
    • Strong coding chops
    • Ability to break down difficult concepts into simple explanations for others
    Bonus points
    • GPU optimization experience
    • Previous industry coding experience

The Perks

Incredibly driven, technical coworkers. A great workplace is people, not happy hours or coconut water, though we have that kind of stuff too:

icon

Run your code on ~10^14 FLOPS of compute set aside solely for experimentation.

icon

Dextro custom GPGPU dev rig(s), with the monitors and peripherals of your choice.

icon

Highly competitive salary and equity package.

icon

Office lunch and dinner made by restaurants in the area.

icon

Healthy snacks, fresh fruit, drinks, bagel brunch Fridays, and an office espresso machine.

icon

Unlimited monthly MetroCard.

icon

100% health, dental, and vision coverage.

icon

Incredible panoramic views of NYC from our 18th floor office in SoHo, with lots of natural light and open space to hang out.

icon

Flexible vacation policy.