Software Development Student , Machine Learning Israel (MLIL) — Firmware

Description

Annapurna Labs builds the custom silicon that powers Amazon's machine learning accelerators. Our Machine Learning Israel (MLIL) Firmware team in Tel Aviv develops the low-level firmware that brings these ML servers to life — from board management and boot flows to device management and health monitoring across the platform.

We are looking for a Software Development Student to join us part-time alongside your studies. You will work on real firmware that ships on production ML servers, writing and testing code, debugging on real hardware, and learning how large-scale accelerator platforms are built from the ground up. You will be working closely with silicon, hardware, and cloud teams to bring new accelerator platforms from first power-on to production. You will be paired with experienced engineers who will mentor you, review your code, and help you grow.

This is a hands-on role for a student who is curious about how hardware and software meet, enjoys solving problems at a low level, and wants to make a real impact on systems used at massive scale.

Key job responsibilities

• Develop, test, and debug firmware components for ML accelerator servers
• Write clean, well-tested code and participate in code reviews
• Investigate and root-cause issues on real hardware in the lab
• Build tools and automation that improve the team's development and validation flow
• Collaborate with firmware, hardware, and platform engineers across the group
  
  

• Currently enrolled in a Bachelor's or higher degree in Computer Science, Computer/Electrical Engineering, or a related field
• At least 3 full semesters remaining until graduation
• Available to work part-time (at least 2-3 days per week) alongside studies
  
  

• Programming experience in C / C++
• Experience with Linux, embedded systems, firmware, or low-level programming
• Exposure to hardware debugging, communication protocols (I2C, PCIe, SPI), or device drivers
• Demonstrated early adopter of AI-assisted development tools- uses LLMs or code-generation agents as part of daily workflow
• Strong problem-solving skills and eagerness to learn in a hardware-focused environment
• Team player, highly motivated, passionate about SW development, and a fast self-learner