Deep Learning Converter Software Developer

Design the tomorrow. Every day.

We are on a mission in taking the world into a new future. Connecting it and enabling new ways of seeing it. By the power of curiosity and passion, using technology, and in full partnership - we lead change day by day. Our vision is built by individuals, connected to one another. We are in the business of creating our own reality, when it’s the daily effort that counts – for an exciting future, one we wish to live in.

About The DNN

The DNN group is a dynamic and innovative team developing solutions for the development, optimization and deployment of AI (DNN)-based applications. We deal with a wide range of activities and technologies, bound tightly together to form our cutting-edge solutions, integrated with Sony's world-leading image sensor products, and the AITRIOS smart-camera platform. Our vision is to continuously push the boundaries of technology and set the bar for advanced-edge AI (DNN) solutions.

The DNN team is looking to hire a motivated and skilled  SW developer. In this role, you’ll have a direct impact on our product. You’ll work on our converter engine that lies at the core of our technology. In a nutshell: we convert neural networks to a mathematically equivalent network that can run on our edge device Our edge device is an AI chip that can infer (not learn, only infer) neural networks at incredible speeds/power-consumption ratios). There are many details naturally, but it is mostly pure programming and modeling, and lib development. It may resemble 2nd year courses algorithm’s parts 1 and 2 from your bachelor’s degree. The converter is written in JVM languages, and it’s built from many small independent parts that fit together. The work involves algorithmic problem-solving and advanced software design skills. We invest much thought in automation for better testing and developer experience. This initiative is driven by and for our developers. The day-to-day work experience varies; sometimes it involves collaborative, team-oriented tasks, while at other times, you may find yourself working independently. Although our converter is cloud native the teams that work on the converter engine write it as an isolated component that is executed in its own container. This position is focused on pure coding and modeling. It has nothing to do with cloud development. Our team is ever-growing, and with that growth comes the opportunity for our team members to evolve as well. We're big on internal development and often see members stepping up to more advanced roles, including leadership positions, as their skills and experience grow. It's about finding the right path for your career, whether that's deepening your technical expertise or exploring managerial responsibilities. We're here to support that journey. This is a challenging position that will be a good fit for strong developers that love solving puzzles and seek a dynamic environment where their creativity in problem-solving and passion for cutting-edge technology can thrive, leading to impactful contributions in the world of deep learning.

An ideal candidate

At our core, we are enthusiastic about programming and believe in making work fun! We're looking for individuals who are as passionate about solving complex puzzles as they are about finding simple and efficient solutions. In our team, the perfect tool for each problem exists - it's about matching the challenge with the right resources and the right developer. Perfection, for us, lies in this synergy of problem, resources, and coder. We welcome a variety of talents and perspectives. Do you see yourself as part of our innovative team?

If so, we would be thrilled to hear from you!

Responsibilities:

• Design and develop key components for our converter.
• Perform neural networks graph manipulation: exploring, rewriting, scheduling, comparing, and debugging.
• Take ownership on projects and features, effectively breaking down tasks into tangible daily goals. 
• Collaborate with other team members both from within and outside the team, a vital aspect of our work culture.
  
  

• Honesty and integrity in all aspects of work
• Capable of autonomous work: a self-starter who requires minimal supervision and can independently drive tasks to completion.
• Strong team player with good personal communication skills.
• Passion for programming!
• At least 5 years R&D experience in Kotlin/Java/Scala (with preference for more extensive experience).
• A bachelor’s degree in computer science or equivalent field.
  
  

• Prior experience in Deep Learning particularly for edge devices
• Proficiency in packaging with Docker/other containers technologies
• Enthusiasm and active contribution to source project both as paid/volunteer development.
• A strong background of Clean, Lean, Immutable, and Test-driven coding principles.