LLM & Multimodal AI Engineer for E‑commerce

TikTok USDS Joint Venture; Seattle (WA); On-site; USD 100,000 - 178,000; Full time

Benefits offered by this job

Health insurance

401(k) matching

Parental leave

Disability insurance

Life insurance

Wellbeing benefits

Paid holidays

Paid sick days

Paid personal time

Job summary

TikTok USDS Joint Venture is seeking a graduate-level technologist to architect and build LLM-based systems that process high-dimensional unstructured data from diverse sources. You will define semantic frameworks and apply advanced techniques such as RLHF and Instruction Tuning to improve performance, while optimizing model serving for production with efficient attention mechanisms.

The role emphasizes high-impact research and production readiness, with onboarding by year-end and opportunities

Qualifications

• Bachelor's/Master's degree in CS, Engineering, Math, Statistics or related discipline.
• Experience in Machine Learning, NLP, or Computer Vision.
• Proficiency in one of C++, Python, Go, or Java.

Responsibilities

• Architect and develop LLM-based systems to process high-dimensional unstructured data.
• Define semantic understanding frameworks with fine-grained intent classification and entity extraction.
• Apply state-of-the-art techniques like Instruction Tuning, RLHF/DPO, and Contextual Prompt Engineering.
• Optimize model serving for production using quantization, distillation, and efficient attention.

Skills

C++

Python

Go

Java

Machine Learning

NLP

Computer Vision

Education

Bachelor's or Master's in CS/Engineering/Math/Statistics

Job description

TikTok USDS Joint Venture is seeking a graduate-level technologist to architect and build LLM-based systems that process high-dimensional unstructured data from diverse sources. You will define semantic frameworks and apply advanced techniques such as RLHF and Instruction Tuning to improve performance, while optimizing model serving for production with efficient attention mechanisms.

The role emphasizes high-impact research and production readiness, with onboarding by year-end and opportunities