Graduate AI/ML Engineer: Multimodal & Agentic AI

TikTok Inc.; San Jose (CA); Sur place; USD 128 000 - 317 000; Plein temps

Résumé du poste

TikTok Inc. seeks PhD graduates or near-completion candidates to develop and optimize LLM/MLLM, Agent, NLP, CV, and recommendation models for a global content ecosystem.

You will design and implement Agentic AI frameworks, integrate multimodal data, and work with researchers, engineers, and business teams to enhance search, discovery, and user experience. Experience with TensorFlow or PyTorch and strong Python/C++ skills are essential.

Qualifications

• PhD completed or near completion in CS/CE/EE/math or related field.
• Strong Python/C++ programming skills.
• Experience with TensorFlow or PyTorch.
• Experience with LLMs/MLLMs, SFT/RL and inference optimization.
• Solid ML fundamentals and AI tool integration.
• Excellent cross-team communication.

Responsabilités

• Develop and optimize LLM/MLLM, Agent, NLP, CV, and recommendation models to improve TikTok's content ecosystem.
• Design, build, and optimize advanced Agentic AI frameworks, focusing on planning, tool use, skills, and memory.
• Implement multimodal AI solutions, integrating video, text, and speech understanding.
• Train and fine-tune deep learning models using TensorFlow, PyTorch, or other ML frameworks.
• Deploy and scale machine learning solutions in a distributed computing environment.
• Work closely with AI researchers, software engineers, and business teams to apply AI technologies effectively.

Connaissances

Python

C++

Deep learning

Communication

Collaboration

Formation

PhD in CS/CE/EE/Math

Outils

TensorFlow

PyTorch

LangChain

LlamaIndex

Post-Training (SFT/RL)

Description du poste

TikTok Inc. seeks PhD graduates or near-completion candidates to develop and optimize LLM/MLLM, Agent, NLP, CV, and recommendation models for a global content ecosystem.

You will design and implement Agentic AI frameworks, integrate multimodal data, and work with researchers, engineers, and business teams to enhance search, discovery, and user experience. Experience with TensorFlow or PyTorch and strong Python/C++ skills are essential.