Responsibilities

The TikTok AI&Visual Search Infrastructure team builds systems that power intelligent search experiences for users around the world. Our infrastructure serves billions of search requests every day and supports products including traditional search, AI-powered search results, in-app chatbots, and visual search. Our work spans both foundational search engine infrastructure and the rapidly evolving AI Search technology stack. We build large-scale indexing and retrieval systems, multimodal search infrastructure, multi-agent and RAG engines, and high-performance LLM/VLM training and serving platforms. As an engineer on the team, you will have the opportunity to work on challenging problems across AI Agents, RAG, LLM/VLM inference/training, distributed systems, search engines and multimodal machine learning. You will collaborate closely with engineers, researchers, and product teams to help shape the next generation of TikTok Search. We embrace a culture of self-direction, intellectual curiosity, openness, and problem-solving. We are looking for talented individuals to join our team. As a graduate, you will get opportunities to pursue bold ideas, tackle complex challenges, and unlock limitless growth. Successful candidates must be able to commit to an onboarding date by the end of the year. Please state your availability and graduation date clearly in your resume. Candidates can apply to a maximum of two positions and will be considered for jobs in the order you apply. The application limit is applicable to our Company and its affiliates' jobs globally. Applications will be reviewed on a rolling basis - we encourage you to apply early. Responsibilities - Design, build, and optimize large-scale infrastructure for TikTok AI Search & Visual Search, supporting billions of online search requests and large-volume offline data processing. - Evolve TikTok’s search agent platform, including multi-agent orchestration, ReAct and tool-calling workflows, RAG indexing & serving engine, and context/memory management. - Deploy and optimize LLM/VLM post-training and inference, including GPU inference acceleration, SFT/RL efficiency, kernel/CUDA optimizations, and offline throughput improvement. - Build and improve search engine components such as offline document processing pipelines, distributed indexing systems for billion/trillion documents , feature extraction services, and multi-stage retrieval/ranking services - Develop scalable data pipelines and distributed systems for processing billion- or trillion-scale text, image, and video datasets, covering offline computation, stream processing, task scheduling, storage, and data quality. - Explore emerging technologies in personalized search, generative search, multimodal understanding, reinforcement learning, model alignment, and evaluation. - Collaborate with machine learning, product, and other engineering teams to deliver accurate, responsive, and engaging AI-powered and visual search experiences.

Qualifications

Minimum Qualifications - Individuals who are completing or have recently completed a Bachelor's degree in Computer Science or a related discipline. - Strong programming skills and a solid foundation in data structures and basic algorithms. - Familiarity with Linux development environment and proficiency in at least one programming language such as C++, Java, Python, or Go - Strong analytical and problem-solving abilities, with an interest in solving complex engineering problems at scale. - Effective communication and collaboration skills, along with a strong sense of ownership and willingness to learn. Preferred Qualifications - Internship, research, course project, or open-source experience in one or more of the following areas: - LLM/VLM serving, RAG, or AI agents - Distributed systems, storage systems, and large-scale data processing - Search engines, information retrieval, recommendation systems, or advertising systems - Understanding of search engine concepts such as query understanding, indexing, retrieval, ranking, and result aggregation - Familiarity with LLM serving frameworks such as vLLM, SGLang, or TensorRT-LLM, or agent development frameworks such as LangGraph, is a plus. - Experience analyzing or optimizing system performance, including latency, throughput, concurrency, memory utilization, or resource efficiency - Ability to learn new technologies quickly and apply them to practical engineering or research problems

Job Information