Senior ML Engineer — LLM & Multimodal Systems Architect

TikTok USDS Joint Venture; San Jose (CA); On-site; USD 187,000 - 360,000; Full time

Benefits offered by this job

Health insurance

401(k) match

Paid parental leave

Disability coverage

Life insurance

Wellbeing benefits

Paid holidays & sick days

Personal time

Job summary

TikTok USDS Joint Venture is seeking an experienced researcher/developer to architect and build LLM-based systems that process large-scale unstructured data from diverse sources. You will define semantic models, implement fine-grained classification, and advance techniques like RLHF and contextual prompting within an e-commerce video and product domain.

The role emphasizes production-ready ML systems and optimized serving, including quantization and distillation, with a focus on multilingual

Qualifications

• MS or PhD in CS, AI, Mathematics, or related field.
• 5+ years in ML, NLP, or Computer Vision.
• Proficiency in C++, Python, Go, or Java.

Responsibilities

• Architect and develop sophisticated LLM-based systems to process high-dimensional unstructured data.
• Define and implement semantic understanding frameworks with fine-grained classification and extraction.
• Apply Instruction Tuning, RLHF/DPO, and Contextual Prompt Engineering techniques.
• Optimize model serving using quantization, distillation, and efficient attention.

Skills

C++

Python

Go

Java

ML/NLP/CV

Education

MS or PhD in CS/AI/Math or related

Job description

TikTok USDS Joint Venture is seeking an experienced researcher/developer to architect and build LLM-based systems that process large-scale unstructured data from diverse sources. You will define semantic models, implement fine-grained classification, and advance techniques like RLHF and contextual prompting within an e-commerce video and product domain.

The role emphasizes production-ready ML systems and optimized serving, including quantization and distillation, with a focus on multilingual