Software Engineering Intern – AI Accelerator Software

Job Description

Technical Intern- Software | DensityAI

The Tone:
This is a software engineering internship at DensityAI, located in California. DensityAI develops the software stack behind custom AI accelerators. This role is crucial for bridging high-level machine learning workloads with specialized silicon, directly impacting performance and efficiency. You will contribute to writing performance-critical code and building essential tooling for our custom AI accelerator.

The TL;DR
• Role: Internship
• Type: Temporary
• Location: California
• Pay: $65–$80 hourly
• Team: Software engineering intern, working alongside kernel, compiler, and architecture teams.
• Mission: Optimize and debug the software layer connecting high-level ML workloads to custom AI accelerator hardware.
• Tech Stack: C/C++, Python

What You’ll Actually Do
• Optimize: Write and optimize compute code for a custom AI accelerator, focusing on tensor operations, data movement, and memory hierarchy.
• Develop: Build and extend profiling and benchmarking infrastructure to measure performance against architectural targets.
• Execute & Debug: Run kernels and workloads end-to-end on the architectural simulator, and help debug unexpected behaviors.
• Automate: Contribute to tooling, test coverage, and automation that supports the daily work of the kernel and compiler teams.
• Collaborate & Document: Take part in code reviews and design discussions with your mentor and the wider team, and write onboarding notes, guides, and documentation.

The Must-Haves
• Background: Currently pursuing a Bachelor of Science in Computer Science, Computer Engineering, Electrical Engineering, or a related field.
• Experience: Demonstrated ability to write and debug real programs beyond coursework snippets.
• Skills: Proficiency in C/C++ and/or Python; strong CS fundamentals including data structures, algorithms, and a working mental model of code execution on a machine; genuine interest in computer architecture and performance, including pipelines and memory hierarchies.
• Bonus: Curiosity, self-direction, and eagerness to learn fast in a startup environment.