Software Engineer - TikTok AI Search Infrastructure

What you'll do

• Evolve the multi-agent LLM engine, including ReAct + Tool calling, DAG-based workflows, and RAG.
• Deploy and optimize text and multimodal LLMs, including inference acceleration, training alignment, and reinforcement learning.
• Build scalable online and offline serving and data platforms for AI Search use cases, and develop personalized AI search capabilities.

What they ask for

• Bachelor or advanced degree in Computer Science or related disciplines.
• At least 5 years of industry experience in recommendation, search, Ads, or ML Infra.
• Strong computer science fundamentals, coding skills, and understanding of algorithms and data structures.
• Experience in at least one of recommendation systems, search engines, advertising engines, Agent/RAG engines, or LLM serving/training.

Nice to have

• Linux performance optimization or large-scale C++ system development.
• Large-scale distributed systems, distributed databases or data processing frameworks, or GPU inference optimization and LLM/VLM serving or training.

Tools

ReAct, Tool calling, DAG-based workflows, RAG, LLMs, VLMs, Linux, C++

AI focus

Develop AI search infrastructure and personalized search experiences; work on multi-agent LLMs, RAG, multimodal model serving and training, inference optimization, model alignment, and reinforcement learning.