Applied Scientist

Amazon Inc.; Bengaluru; On-site; INR 1,800,000 - 3,600,000; Full time

Job summary

Amazon Science in Bengaluru is seeking a scientist to help turn ideas into scalable models for Kindle Direct Publishing. You will work on multi-agent systems, multimodal models for catalog/reader experience, and post-training/fine-tuning of models for deployment.

You will collaborate with global teams, design rigorous experiments, and document insights to drive product impact and policy alignment across languages and regions.

Qualifications

• PhD or MSc in CS/CE/ML
• 3+ years of building machine learning models for business applications
• Experience programming in Java, C++, Python or related language

Responsibilities

• Build the multi-agent system behind KDP's author support assistant across agents and tools; personalize and evaluate for continuous improvement.
• Build multimodal models of book attributes for similar-book search, catalog quality, and category/keyword recommendations.
• Post-train open models on KDP expertise and deploy them where they beat baselines on accuracy, cost, or speed.
• Collaborate with scientists in the US and engineering partners; design offline/online experiments and write readouts with recommendations.

Education

PhD or MSc in CS/CE/ML

Job description

Description

Help millions of authors around the world get their books published, discovered, and supported. As one of the first scientists on our new team in Bengaluru, you'll help turn ideas into models authors rely on every day, building agentic solutions with large language models (LLMs), multimodal models, and fine-tuning.

Key job responsibilities

• Build the multi-agent system behind KDP's author support assistant: orchestration across specialized agents and tools, personalization, and the evaluation that drives its continuous improvement.
• Build multimodal models of book attributes, and use them for similar-book search across languages, catalog quality, and category and keyword recommendations.
• Post-train open models on KDP expertise (for example with supervised fine-tuning and reinforcement learning), and deploy them where they beat general-purpose models on accuracy, cost, or speed.
• Work closely with our scientists in the US and with engineering partners to take your work to production, design rigorous offline and online experiments, and write clear readouts with recommendations.

About The Team

Kindle Direct Publishing (KDP) lets anyone publish a book. We are the science organization for Amazon's Books business, now building our science teams in Bengaluru. Our team works in three connected areas: a support assistant that answers authors' questions from their own account data, in every language KDP supports; multimodal models that use book attributes to keep the catalog trustworthy and help authors reach the right readers; and fine-tuned models that power the assistant and catalog models with KDP expertise. We serve authors, from first-time writers to full-time publishers. You will work closely with engineering, product, and policy partners. We value safety before speed and shared foundations over one-off fixes. Scientists here get dedicated time for innovation, which can lead to patents and conference papers.

Basic Qualifications

• PhD, or a Master's degree and experience in CS, CE, ML or related field
• 3+ years of building machine learning models for business application experience
• Experience programming in Java, C++, Python or related language

Preferred Qualifications

• Experience building LLM-based agents or multimodal generative AI workflows, including tool use, retrieval-augmented generation, or evaluation
• Experience fine-tuning or post-training models, including with reinforcement learning
• Experience with large-scale retrieval, embeddings, or ranking, including text and image embeddings
• Experience with multilingual natural language processing or computer vision
• Experience with SQL and large-scale data processing
• Experience in patents or publications at top-tier peer-reviewed conferences or journals
• Demonstrated experience leveraging generative AI tools to enhance workflow efficiency and productivity, with the ability to craft effective prompts and critically evaluate AI-generated outputs in a professional setting

Our inclusive culture empowers Amazonians to deliver the best results for our customers. If you have a disability and need a workplace accommodation or adjustment during the application and hiring process, including support for the interview or onboarding process, please visit https://amazon.jobs/content/en/how-we-hire/accommodations for more information. If the country/region you’re applying in isn’t listed, please contact your Recruiting Partner.