Job Description
Amazon Devices is hiring an applied scientist for generative AI on edge devices in Bengaluru. The SSG Science team builds and optimises next-generation edge models for products such as Kindle, Fire tablets, Fire TV, Echo and Astro, co-designing them with Amazon’s custom ML hardware.
At a glance
- Role: Applied Scientist, SSG Science (Amazon Devices)
- Location: Bengaluru, Karnataka, India
- Posted: 2 September 2026
- Experience: 2+ years building ML models for business use; PhD, or a master’s degree plus 2+ years of applied research
- Job ID: 10526535
- Apply: open until filled (re-checked 13 October 2026)
About the generative AI on edge devices role
- Quantise, prune, distil and fine-tune generative AI models for edge platforms
- Understand Amazon’s Neural Edge Engine deeply and invent new optimisation techniques for it
- Analyse deep learning workloads and advise on mapping them to the engine
- Draw on information theory, scientific computing, deep learning theory and non-equilibrium thermodynamics
- Train custom generative models that beat the state of the art and lead to production models
- Work with compiler engineers, hardware architects and product teams; publish and present at venues such as NeurIPS, ICLR and MLSys
What Amazon is looking for
- 2+ years building ML models for business applications
- A PhD, or a master’s degree and 2+ years of applied research experience
- Patents or publications at top peer-reviewed venues
- Programming in Java, C++, Python or similar
- Strong maths foundations, including linear algebra and calculus
Nice to have
- Professional software development
- Experience taking ML models through the full production lifecycle
Pay and location
Amazon’s posting does not state a salary; pay is discussed during the hiring process. The role is based in Bengaluru, India.
How to apply
Apply through the official Amazon job posting. Amazon’s posting does not give a closing date, so the role is open until filled; ResearchJobs.in will re-check this listing on 13 October 2026. Amazon lists the minimum and preferred qualifications on the posting; read both before applying. 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.
Hiring institution: Amazon
Official advertisement: www.amazon.jobs
How to prepare for this application
- Compression techniques: revise post-training quantisation, quantisation-aware training, pruning and distillation for transformers.
- Hardware awareness: be ready to reason about memory bandwidth, operator fusion and why a model is slow on an accelerator.
- Diffusion and LLMs on device: know recent work on running generative models on phones and edge chips.
- Theory: the posting mentions information theory and thermodynamics; know where they appear in diffusion models.
- Leadership Principles: prepare STAR stories.
About Amazon in India
Amazon runs large science and engineering teams in India, including in Bengaluru, Hyderabad, Chennai and Gurugram. Applied scientists work on machine learning problems across retail, logistics, advertising, devices and cloud services.