Machine Learning Engineer (Ads Core) - User Experience

Description

Ads Core team is chartered to build key monetization components across various ad delivery stages:
1. We build up ranking, bidding, budget, format, diagnosis and other frameworks that serve as a mid platform to enable other ad teams to iterate their products in parallel.
2. We implement outstanding traffic strategies to maximize revenue under the constraint of user experience and achieve complete exploration of advertiser's audience.
3. Our model driven automation solutions optimize ad delivery performance from end to end.

Responsibilities:
1. Participate in and take charge of algorithm optimization for user ad experience in TikTok's global business, optimize the conversion efficiency between ad revenue and user experience by intervening in the ad bidding and ranking logic via signals, and realize the sustainable and sound development of the ad ecosystem;
2. Responsible for the end-to-end technical iteration of advertising experience optimization, covering three core modules: sorting upgrade, duplication governance, and crowd strategy scheduling, and continuously improve the collaborative efficiency of experience and monetization through underlying model infrastructure and strategy innovation;
3. Deeply explore user behavior feedback and experience pain points, and systematically solve core user experience issues through end-to-end technology upgrade, covering areas including engagement model optimization, implementation of personalized frequency distribution, and upgrade of user sensitivity modeling;
4. Lead the underlying capability building based on LLM, implement user intent recognition and semantic understanding technology for advertising content, integrate multimodal representation capabilities with the large model as the core, and provide core capability support for end-to-end experience optimization strategies.

Requirements

Minimum Qualifications:
- Possess excellent coding skills, solid foundation in data structures and algorithms, as well as solid theoretical knowledge and practical experience in machine learning and deep learning;
- Be familiar with at least one mainstream deep learning programming framework (TensorFlow/PyTorch), and have a good command of its underlying architecture and implementation mechanism;
- Excellent ability to analyze and solve problems, with a strong passion for tackling challenging issues;
- Have a passion for technology, as well as excellent communication skills and team spirit.

Preferred Qualifications:
- Have work experience related to models or customer ecosystem in advertising and recommendation scenarios;
- Have published papers at relevant conferences such as ICML, NIPS, ICLR, RecSys, KDD, etc. ;
- Experience in open-source communities of deep learning programming frameworks.