Neural Network Optimization Engineer

Untether AI is building the world’s highest performance pure-digital AI inference chip. We’re a rapidly growing Toronto-based startup, with employees across Canada and the US, building next generation hardware AI accelerators for neural net inference. We are investing in software in a big way in order to make it as seamless as possible for researchers and developers to successfully deploy neural networks on our hardware. We’re looking for best in class engineers to join our existing top-notch team. When you join Untether AI, you will be part of a team that designs, develops and verifies the software that interacts with our chip, collaborating with our hardware engineers and with fellow software engineers in the process. By creating software that fully realizes the capabilities of the hardware, you will help get AI inference to the general populace.As part of this exceptional team, you are able to - and get excited by - identifying functional/performance bottlenecks and how to alleviate them in order to achieve scalable and reliable software. You excel in an environment with complex software and hardware designs. The Neural Network Optimization Engineer addresses the design, development and optimization of leading edge neural networks for our customers. The role requires hands-on development of neural networks, involves working with a cross-functional team of engineers that are focused on customers, and acts as a conduit between our customers and our software and hardware teams. By using and/or developing our software tools (kernel support, physical allocation placement, etc.), in particular for customer neural nets that require functionality/performance beyond what's available in our standard software releases, they ensure the most effective use of our technology by our customers.The Neural Network Optimization Engineer understands our products, technology and applications and is an expert in developing neural networks for the world’s highest performance AI accelerator chips. They will diagnose and address the challenges of real-world uses, resolve them and escalate to appropriate teams as needed. They will also provide useful feedback from these experiences to our engineering teams to improve existing products and to the CTO/architecture team for input on next gen products.ResponsibilitiesUse and advance our products in the development and delivery of leading edge customer neural networks and applicationsUnderstand and evaluate neural network performance at a chip, system and application level through performance profiling and benchmarkingIdentify/diagnose challenges in using our products for specific use cases and define and develop projects to resolve them. Engage with the Customer Neworks Lead on any escalations to the software and hardware teamActively engage with customers to ensure they feel supported by Untether AI post sale and delivery of the productPartner with the Products team to ensure they are up to date on client experiences and their feedback so that interactions with customers appear seamlessManagement of program / project teams to deliver Neural Network Model Garden Models including the identification, design and implementation of C++ low-level flexible programs (kernels) for various neural net operationsWhen required development of complexed kernels and compiler strategies to address the network requirementsCommunicate performance optimization ideas both to compiler/kernel engineers and to architects working on future product generationsPreferred skills & ExperienceEnthusiastic and knowledgeable about all things “AI”Computer Science, engineering, or related degree At least 5 years of software development experience, specifically in AIKnowledge of Neural Network basic operator algorithms - Convolutions, Transformers, RNNsExperience with end to end software development, specifically data structures and algorithms, experience with software architecture patterns, etc.Experience working in cross-functional teamsStrong bias for teamwork and effective problem solvingExperience with tuning and optimizing code and Neural Networks for high performanceStrong communication skill and a background in program / project managementSuperior problem solving skills, both technical and interpersonalThinking about applying? We’re a pretty welcoming bunch of people. If we’ve piqued your interest, you’re passionate about the same things we are but you aren’t sure if you check all the boxes, please apply anyway. We’re a great place to work, an even better place to learn and we focus on both capability and potential!Besides being a pretty friendly group, we’re also pretty diverse - ethnically, experientially and in our physical and mental abilities. If you are someone who needs accommodation during the interview process and/or in the workplace, please send a note along with your resume to [email protected] and we’ll be in touch.What are some of the perks that you will receive being part of Untether AI? Aligned to Untether’s philosophy, our employees enjoy the same perks, regardless of role or level. In part, these include:20 vacation daysStrong health and extended health benefitsUnlimited sick daysStock optionsBuilding chips and software that will change the worldA little bit more about Untether AIUntether AI has developed a groundbreaking new architecture that brings neural net inference to new levels of performance and efficiency. We’ve already sold our product to smart clients who want to get in at ground zero. We’ve done this while continuing to improve our technology creating ultra-efficient, high performance AI chips that eliminates the data movement bottleneck that costs energy and performance in traditional architectures. We’re a team made up of scientists, engineers and entrepreneurs and have the support of tier one investors. We recently received $125 million in our series B funding round which enables us to expand our customer engagements, enhance our software offering, and build the next generation of industry leading AI inference products. Join us to be part of something big - a chance to create the future of AI.