Applied Scientist I, Ads Trust Science
Amazon
41 minutes ago
Full-time
On-site
Bengaluru, Karnataka, India
Applied Scientist I, Ads Trust Science, Bangalore
Every ad that reaches a customer through Amazon DSP has passed through a system you'll help build. The Ads Trust Science team protects one of the world's largest advertising ecosystems , We moderate millions of ad requests a day across 1P and 3P publisher platforms globally across different languages. And we're scaling by an order of magnitude from here.
This isn't a research team working in isolation, it's a science team whose models ship into production and directly decide what customers should not see. You'll build ML models that understand ads the way a human reviewer would: reading text, parsing images, catching intent, across languages and cultures. You'll work at the frontier of multimodal content understanding, combining vision and language models to solve problems that don't have off-the-shelf solutions, because the scale and stakes are unlike almost anywhere else.
What you'll actually do:
- Design and train multimodal (vision + language) ML models that flag or clear ads at massive scale
- Push these models beyond English. You will build systems that generalize across languages, scripts, and cultural context
- Take models from notebook to production: write the code, build the pipelines, and own the systems that moderate millions of ads a day
- Partner closely with engineers and fellow scientists to turn a research idea into something that runs reliably at Amazon scale
- See your work matter immediately: a model you ship this quarter is protecting customers next quarter
Why this role:
- Rare combination: real research problems (open-vocabulary understanding, low-resource languages, cross-modal reasoning) with real production impact and real scale
- You won't be the only scientist working on a narrow slice rather you'll have end-to-end ownership from problem formulation to deployment
- Trust and Safety at Amazon's size is a genuinely hard, underexplored ML problem. Most of what you'll build doesn't exist in a textbook yet
Basic Qualifications: - Experience building machine learning models or developing algorithms for business application
- Experience in state-of-the-art deep learning models architecture design and deep learning training and optimization and model pruning
- Currently has, or is in the process of obtaining, a Master's degree or equivalent in Engineering, Computer Science, Machine Learning, Operations Research, Statistics, or related fields
- Experience researching about machine learning, deep learning, NLP, computer vision, data science
- Experience programming in Java, C++, Python or related language
- Experience with SQL and an RDBMS (e.g., Oracle) or Data Warehouse Preferred Qualifications: - Experience applying theoretical models in an applied environment
- Have publications at top-tier peer-reviewed conferences or journals
- Master's degree
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.
Every ad that reaches a customer through Amazon DSP has passed through a system you'll help build. The Ads Trust Science team protects one of the world's largest advertising ecosystems , We moderate millions of ad requests a day across 1P and 3P publisher platforms globally across different languages. And we're scaling by an order of magnitude from here.
This isn't a research team working in isolation, it's a science team whose models ship into production and directly decide what customers should not see. You'll build ML models that understand ads the way a human reviewer would: reading text, parsing images, catching intent, across languages and cultures. You'll work at the frontier of multimodal content understanding, combining vision and language models to solve problems that don't have off-the-shelf solutions, because the scale and stakes are unlike almost anywhere else.
What you'll actually do:
- Design and train multimodal (vision + language) ML models that flag or clear ads at massive scale
- Push these models beyond English. You will build systems that generalize across languages, scripts, and cultural context
- Take models from notebook to production: write the code, build the pipelines, and own the systems that moderate millions of ads a day
- Partner closely with engineers and fellow scientists to turn a research idea into something that runs reliably at Amazon scale
- See your work matter immediately: a model you ship this quarter is protecting customers next quarter
Why this role:
- Rare combination: real research problems (open-vocabulary understanding, low-resource languages, cross-modal reasoning) with real production impact and real scale
- You won't be the only scientist working on a narrow slice rather you'll have end-to-end ownership from problem formulation to deployment
- Trust and Safety at Amazon's size is a genuinely hard, underexplored ML problem. Most of what you'll build doesn't exist in a textbook yet
Basic Qualifications: - Experience building machine learning models or developing algorithms for business application
- Experience in state-of-the-art deep learning models architecture design and deep learning training and optimization and model pruning
- Currently has, or is in the process of obtaining, a Master's degree or equivalent in Engineering, Computer Science, Machine Learning, Operations Research, Statistics, or related fields
- Experience researching about machine learning, deep learning, NLP, computer vision, data science
- Experience programming in Java, C++, Python or related language
- Experience with SQL and an RDBMS (e.g., Oracle) or Data Warehouse Preferred Qualifications: - Experience applying theoretical models in an applied environment
- Have publications at top-tier peer-reviewed conferences or journals
- Master's degree
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.