Artificial Intelligence Researcher

AMD; Bengaluru; On-site; INR 1,800,000 - 2,400,000; Full time

Job summary

AMD Bengaluru's AI Models team is seeking exceptional ML scientists and engineers to advance training and inference for LLMs, LMMs, and foundation models on AMD accelerators. You will work on scalable pre-training, instruction tuning, alignment, and optimization, collaborating with a world‑class research group to push the boundaries of performance and efficiency.

This role offers opportunities to publish, contribute to open source, and influence the strategy of self-evolving agents and agentic

Qualifications

• Advanced degree (Master’s or PhD) in machine learning, computer science, artificial intelligence, or a related field is expected.
• Exceptional Bachelor’s degree candidates with years of relevant research experience may also be considered.

Responsibilities

• Improve upon the state-of-the-art in Generative AI model architectures and their compatibility on AMD accelerators.
• Accelerate the training and inference speed through various optimisations.
• Build AI agents to automatically write and evaluate efficient kernels and code
• Drive continuous improvement of infrastructure and development ecosystem
• Publish your research at top‑tier conferences, workshops and/or through technical blogs.
• Engage with academia and open‑source ML communities.

Skills

Python

PyTorch

TensorFlow

CUDA/HIP/Triton

Transformers

AI agents

Kernels

Distributed training

Education

PhD in ML/CS/AI or related field

Master’s in ML/CS/AI or related field

Tools

HIP

CUDA

Triton

Job description

The AI Models team is looking for exceptional machine learning scientists and engineers to explore and innovate on training and inference techniques for large language models (LLMs), large multimodal models (LMMs), image/video generation and other foundation models as well as self-evolving agents on top of these. You will be part of a world-class research and development team focussing on efficient and scalable pre-training, instruction tuning, alignment and optimization. As an early member of the team, you can help us shape the direction and strategy to fulfil this important charter.

THE PERSON:

This role is for you if you are passionate about reading through the latest literature, coming up with novel ideas, and implementing those through high quality code to push the boundaries on scale and performance. The ideal candidate will have both theoretical expertise and hands‑on experience with developing and optimizing LLMs, LMMs, and/or diffusion models.

KEY RESPONSIBILITIES:

• Improve upon the state-of-the-art in Generative AI model architectures and their compatibility on AMD accelerators
• Accelerate the training and inference speed through various optimisations.
• Build AI agents to automatically write and evaluate efficient kernels and code
• Drive continuous improvement of infrastructure and development ecosystem
• Publish your research at top‑tier conferences, workshops and/or through technical blogs.
• Engage with academia and open‑source ML communities.

PREFERRED EXPERIENCE:

• Strong development and debugging skills in Python.
• Experience in deep learning frameworks (like PyTorch or TensorFlow), distributed training tools as well as various agentic AI frameworks
• Solid understanding of various types of transformers and state space models.
• Experience in writing kernels using HIP, CUDA, Triton, etc.
• prior experience in building self‑evolving AI agents for various challenging tasks (like code generation, discovery, etc.)
• knowledge of latest research in the field of AI agents
• familiar with techniques to do hardware‑efficient training and inference of large language models and also familiarity with existing frameworks (e.g. vllm, sglang, etc.)
• expertise in model architecture (beyond just transformers only models) and latest research in this field
• having relevant open‑source contributions
• Strong publication record in top‑tier conferences, workshops or journals.
• Solid communication and problem‑solving skills.
• Passionate about learning new stuff in this domain as well as innovating on top of it

ACADEMIC CREDENTIALS:

• Advanced degree (Master’s or PhD) in machine learning, computer science, artificial intelligence, or a related field is expected. Exceptional Bachelor’s degree candidates with years of relevant research experience may also be considered.