Senior Algorithm Engineer - TikTok Location Based Services (Singapore)

Description

TikTok LBS Platform's goal is to build a global high-quality Location Based Service database and high-performance LBS services that connects TikTok users with nearby merchants and local services.

- Responsible for the design and development of algorithms related to the in-house construction of POI (Point of Interest) data for TikTok Local Services, including but not limited to data mining, quality evaluation, consistency validation, and intelligent review systems.
- Design and implement core Agent modules, including tool invocation planning, memory management, skill orchestration, and retrieval mechanisms; focus on self-healing capabilities, fallback strategies, and edge-case handling.
- Build the Agent evaluation framework by defining success metrics, designing regression testing methodologies for non-deterministic outputs, and quantifying trade-offs across cost, latency, and quality.
- Drive the application of large language models in POI scenarios, including fine-tuning, model distillation, knowledge base construction, and retrieval systems, with a strong focus on improving POI data quality.
- Deliver high-efficiency engineering outcomes through AI Coding workflows, leveraging AI toolchains to expand development capabilities and independently complete end-to-end delivery from algorithm design to deployment.
- Develop a deep understanding of POI data scenarios, identify real-world problems through data insights, and continuously drive algorithm iteration and optimization.

Requirements

Minimum Qualifications:
- Strong problem diagnosis and troubleshooting skills: able to systematically break down issues, generate hypotheses, and validate them quickly when facing problems such as abnormal model loss or unexpected Agent behaviors, rather than relying on trial-and-error solution switching.
- Solid system design mindset: not only focused on model performance, but also highly attentive to system stability, including fallback strategies, fault tolerance, and edge-case handling.
- Strong programming fundamentals: proficient in data structures and algorithms, with the ability to write code that is readable, debuggable, and maintainable.
- Fundamental model iteration capabilities: deep understanding and hands-on experience in one or more of the following areas — data mining, deep learning, computer vision (CV), large language models (LLM), or masked language models (MLM).

Preferred Qualifications
- Hands-on experience in large language model fine-tuning and preference learning.
- Experience in designing or developing Agent, Skill, and Memory related modules.
- Production-level engineering experience with non-deterministic systems, such as recommendation, search, or content understanding systems.
- Strong practical experience with AI Coding workflows, including end-to-end cross-domain project delivery.