Bilingual Machine Learning Engineer, Content Understanding/Recommendation
- Tokyo
- Partial Remote
- Full-time
- September 10, 2024
About SmartNews
SmartNews is a global leader in redefining information and news discovery, committed to providing users with accurate, timely information and supporting quality journalism. We combine the power of our unique machine-learning technology with the expertise of our first-rate editorial team to bring news that matters to millions of users from our over 3,000 global publishing partners.
Founded in 2012 in Tokyo, Japan, SmartNews has offices in Tokyo, San Francisco, Palo Alto, New York, and Singapore.
The Team
We are currently hiring for our Content Understanding and Recommendation Algorithm Team.
Content Understanding Team
Content understanding team is to advance the capabilities of content analysis through innovative and cutting-edge technologies. We strive to transform how content is understood, categorized, and utilized across various platforms, ensuring accuracy, relevance, and quality. By leveraging advanced machine learning, NLP, and data analytics, we aim to support and enhance downstream applications such as search, recommendations, and ads. Our goal is to deliver meaningful insights and improve user engagement by providing precise and contextually aware content analysis solutions. Through continuous research, development, and collaboration, aiming to make a significant impact on the content ecosystem for SmartNews.
Recommendation Algorithm Team
The Recommendation Algorithm Team serves as the architect behind the cutting-edge algorithms and systems that power the SmartNews Product. Our scope encompasses a range of surfaces, most notably For You, Follow Recommendations, Search and so on..
At the core of the Recommendation Algorithm Team’s mission is the commitment to provide SmartNews users with an unparalleled experience in discovering quality information. Through our intricate algorithms, we curate and present content in a manner that resonates with personalization, fostering engagement and, in turn, bolstering critical business metrics including DAU, revenue, and user engagement.
Content Understanding & Recommendation Algorithm Team
Both teams thrive on the challenge of transforming vast data into actionable insights. We continuously refine and innovate upon our algorithms to ensure that the information presented is not only relevant but also delivered with precision, enriching the user journey and driving SmartNews' competitive edge.
In collaboration with cross-functional teams, we contribute to the seamless integration of our algorithms into various product facets, each designed to enhance user interactions and overall satisfaction. By blending data science, engineering prowess, and a passion for innovation, our team is committed to shaping SmartNews into a trusted and indispensable source of personalized content for millions of users worldwide.
The Role
In the Content Understanding role, you will leverage advanced content analysis and machine learning techniques to enhance the understanding and classification of diverse content types. Your mission is to drive innovation in content categorization, fake news detection, and bias analysis, ultimately supporting downstream applications such as personalized product recommendations.
In the Recommendation Algorithm role, you will utilize the power of advanced machine learning techniques, these teams not only optimize content distribution and user engagement but also provide the essential foundation for continuous product innovation and overall business success.
Responsibilities
- Common Responsibilities Between Both Teams:
- Research and develop cutting-edge algorithms.
- Work end-to-end, from data and model to service.
- Manage expectations between product side and ML limitation.
- Develop ML pipeline to make engineers more productive.
- Content Understanding Team:
- Analyze and categorize content to extract key insights, themes, and trends, ensuring alignment with user needs and product goals.
- Generate actionable signals from content to enhance recommendation systems, driving personalized and relevant user experiences.
- Collaborate with product teams to integrate content insights into product features and strategies, contributing to continuous improvement and product success.
- Recommendation Algorithm Team:
- Innovate and implement machine learning algorithms, seamlessly integrating rule-based optimization strategies, resulting in tangible enhancements across product metrics for our diverse array of offerings such as ads, feed, and follow recommendations.
- Design, develop, and iterate recommendation and ranking algorithms on predictive models for candidate generation and ranking, including but not limited to building real-time data pipelines, feature engineering, model optimization and innovation.
Requirements
Minimum requirements
- 1+ years (0 for new grad) of experience in designing and implementing machine learning algorithms, and applying them to real world problems
- Solid Machine Learning background and deep understanding of certain domain of machine learning techniques, especially in recommendation systems, natural language processing, computer visions
- Strong software development skills with proven record of shipping changes to production that improved product metrics with machine learning technologies
- Able to have deep end-to-end understanding of sophisticated ranking systems and can proactively detect problems and make improvement suggestions
- Good written and spoken communication skills, can work across functional teams
- Expert coding abilities in multiple programming languages (e.g. Java, C++, Python, Scala)
- MS or BS in computer science, mathematics, physics or other quantitative fields
- Fluent in Japanese
Nice to have
- Native in Japanese
- Experience relevant to search, recommendation, and Ads technologies.
- Experience with cloud-based architecture, such as Amazon Web Services (AWS).
- Strong interest in news and ads content understanding, aligned with our mission.
- In-depth knowledge of Generative AI (GAI) and expertise in leveraging it effectively.
Working condition
- Office Location: Tokyo
Benefits
- Annual health check covered by the company
- Visa sponsorship and overseas relocation support available for eligible candidates
About SmartNews
Recently crowned Japan's latest tech unicorn, this company runs a wildly successful news service that’s powered by machine learning. It's an engineering-focused company (One of co-founders is an engineer) and they have offices in San Francisco, Palo Alto, Shanghai, Beijing that are apparently doing quite well too.
The mission of SmartNews is to discover and deliver quality stories to the world. Since its inception in 2012, the SmartNews app has been downloaded over 50 million downloads and maintained a monthly active user count of over 20 million.
In addition to being one of the most popular smartphone apps in Japan, SmartNews has worked hard on building their business in the US. They now have partnerships with over 300 major news publications in the US like CBS, Entertainment Weekly, Business Insider, National Geographic and more. They’ve had a great deal of success due to their ML-based sorting and recommendation systems.
SmartNews has a highly diverse development team with many world-class engineers and others working from their offices in Japan,China and the US. They maintain a very strict hiring bar for technical hires and an engineering-driven culture that makes them one of the strongest tech companies in Japan when it comes to technical skill level.
SmartNews is looking to hire the best people regardless of background, so they’re willing to hire people that don’t speak Japanese and (under normal circumstances) bring them to Japan from overseas if necessary. Everyone works together as one team regardless of location, and SmartNews provides everything their employees need to do their best work and maintains a Silicon Valley-inspired culture for engineers in Tokyo.
Get Job Alerts
Sign up for our newsletter to get hand-picked tech jobs in Japan – straight to your inbox.