Vellex Computing

AI Systems Architect

Vellex Computing

full-time

Posted on:

Location Type: Hybrid

Location: Delhi • 🇮🇳 India

Visit company website
✨ AI Apply
Apply

Job Level

Mid-LevelSenior

Tech Stack

PyTorchTensorflow

About the role

  • Analyze state-of-the-art AI hardware to identify fundamental bottlenecks in performance and energy efficiency.
  • Design and simulate hybrid hardware-software architectures that leverage the best of both digital and analog computing.
  • Develop proof-of-concept firmware and software to validate architectural decisions on embedded platforms.
  • Collaborate with analog ASIC designers to define the requirements and interfaces for future hardware generations.

Requirements

  • Education: MS or PhD in Computer Engineering, Electrical Engineering, or Computer Science. We value relevant academic research or industry experience in computer architecture, hardware-software co-design, or embedded machine learning.
  • Technical Skills: Deep understanding of computer architecture (memory hierarchies, data flow in GPUs/NPUs) combined with a solid grasp of ML mathematics (training, backpropagation, optimization). Strong proficiency in C/C++ for embedded environments and defining hardware interfaces is essential.
  • Tool Proficiency: Experience with standard ML frameworks (e.g., PyTorch, TensorFlow) and performance profiling tools.
Benefits
  • Compensation: Competitive salary and equity package.
  • Flexible Work: Hybrid model based in Delhi with flexible hours to accommodate international collaboration.
  • First Principles Thinking: We love physics. We believe that by breaking problems down to their fundamental truths, even the most complex challenges can be solved.
  • Share and Support: Our culture is collaborative. We are all about supporting each other through the ups and downs of startup life.
  • Adapt and Excel: Agility is part of our DNA. We view mistakes as learning opportunities.

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard skills
computer architecturehardware-software co-designembedded machine learningC/C++ML mathematicstrainingbackpropagationoptimizationperformance profiling