Applied Research Scientist, AI Agents and Models at AMD (Bangalore)

Job Description

AMD is hiring an Applied Research Scientist for AI agents and models in Bangalore. The role sits in its AI Models team, which researches faster and more scalable training and inference for large language models, large multimodal models, image and video generators and other foundation models, and builds self-evolving agents on top of them. The team’s focus is efficient pre-training, instruction tuning, alignment and optimisation, and AMD says early members will help set its direction.

At a glance

• Position: Applied Research Scientist, AI Agents & Models (requisition 88801)
• Where: Bangalore, India
• Team: AMD AI Models team
• Degree: Master’s or PhD expected
• Posted: 21 August 2026
• Apply: open until filled (re-checked 15 October 2026)

What the applied research scientist does in AMD’s AI agents and models team

• Speed up the training and inference of AI models and applications with efficiency techniques
• Build AI agents that automatically write and evaluate efficient GPU kernels and workload-optimisation code
• Keep improving the team’s infrastructure and development ecosystem
• Publish research at top-tier conferences and workshops, or through technical blogs
• Work with academia and open-source machine learning communities

Why this role stands out

Using LLM agents to write and tune GPU kernels links frontier AI research directly to hardware performance. AMD designs the Instinct accelerators used for AI training and inference, so progress on kernel-writing agents and efficient training feeds straight into how well models run on its chips. The posting also expects publications and open-source engagement.

What AMD is looking for

• A Master’s or PhD in machine learning, computer science, artificial intelligence or a related field; exceptional Bachelor’s graduates with years of relevant research experience may also be considered
• Theoretical knowledge and hands-on experience of developing and optimising LLMs, large multimodal models and/or diffusion models
• Strong Python development and debugging skills
• Experience with deep learning frameworks such as PyTorch or TensorFlow, distributed training tools and agentic AI frameworks
• A solid grasp of model architectures beyond transformers, such as state-space, diffusion and flow models
• Good communication and problem-solving skills, and a habit of reading the latest literature and turning ideas into high-quality code

Nice to have

• Writing and optimising GPU kernels in HIP, CUDA or Triton
• Building self-evolving AI agents for tasks such as code generation or discovery
• Hardware-efficient training and inference of generative models, and serving frameworks such as vLLM and SGLang
• Open-source contributions and a strong publication record at top venues

Pay and location

AMD’s posting does not give a salary; it links to a summary of AMD benefits. The role is full time in Bangalore, and AMD states that the posting is for an existing vacancy.

How to apply

Apply through the official AMD job posting. AMD’s posting does not give a closing date, so the role is open until filled; ResearchJobs.in will re-check this listing on 15 October 2026. AMD says it may use AI to help screen applicants, and it does not accept unsolicited CVs from recruitment agencies. Check the posting for location and work-authorisation details before you apply.

See all our artificial intelligence jobs, or browse more research jobs on ResearchJobs.in.