Bengaluru · On-site

AI Software Engineer - AI Frameworks

Full-timePosted 6 Nov

Salary: Not disclosed
Experience: 6–12 years
Eligible batches: 2020
Degree: B.Tech, M.S, M.Tech
Location: BengaluruOn-site

About the role

Intel's Data Center and AI group is looking for an AI Software Engineer to design and optimise high‑performance AI serving and training frameworks such as SGLang, vLLM and PyTorch. The role focuses on GPU kernel development, scheduling, memory management and overall framework optimisation for Intel accelerators and next‑generation GPUs.

Responsibilities

Create custom GPU kernels for attention, collective ops and other primitives to boost performance on Intel hardware.
Enhance framework components – scheduling, memory handling and execution pipelines – to improve throughput and latency.
Integrate high‑throughput kernels using GPU programming frameworks and validate their impact.
Identify optimisation opportunities across the AI framework stack for production workloads.
Collaborate with open‑source communities (SGLang, vLLM, PyTorch) and contribute upstream.
Debug and optimise complex multi‑layered systems involving kernels, memory hierarchies and distributed AI scenarios.

Free interview prep

Prepare for the interview

All lists

Picked for this role's title and skills: the questions that come up most, each with a short answer to say out loud.

Machine learning interview questions40 questions · ≈ 18 min to revise

Asked most often · 19 of 40

Python interview questions40 questions · Data structures, Functions, OOP

Eligibility

Bachelor’s (B.Tech) or Master’s (M.S./M.Tech) in Computer Science, ECE or related discipline.
6‑12 years of professional experience in AI software engineering.
Strong communication skills and ability to work with globally distributed teams.

Skills

Advanced C++ (C++14/17) and solid Python knowledge.
Hands‑on experience writing and optimising GPU kernels (CUDA, CUTLASS, Triton).
Deep familiarity with AI frameworks such as SGLang, vLLM, PyTorch.
Expertise in scheduling, memory management and performance profiling for AI workloads.
Understanding of GPU architecture, memory hierarchies and hardware‑software co‑optimisation.
Experience with distributed AI execution and multi‑GPU scaling.

Resume keywords

Terms from this listing. Many screening tools match exact words, so mirror the ones you genuinely have.

Asked for

C++
Python
GPU
Parallel programming
AI frameworks
Cuda
Cutlass
Triton
Performance profiling
Distributed computing

Nice to have

Compiler algorithms
Quantization

Tailor my resume with these

Benefits

Opportunity to work on cutting‑edge AI accelerator technology.
Collaboration with leading open‑source AI projects.
Exposure to high‑impact production systems at a global semiconductor leader.

Apply now

You'll apply on intel.wd1.myworkdayjobs.com

Find a referralPeople at Intel on LinkedIn

Apply link checked by CareersAt.Tech and re-tested nightly. You never pay to apply. Report listing

About Intel

Designs and makes processors and other chips for PCs, data centres and edge devices, and manufactures chips for other companies through Intel Foundry.

Founded 1968 · 5000+ employees · HQ Santa Clara, California, United States

More about Intel

Similar roles

More backend jobs