Conversational AI Research Scientist - PhD Graduate

TikTok; San Jose (CA); Presencial; USD 162.000 - 388.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Medical, dental, vision insurance

401(k) with company match

Paid parental leave

Short-term and long-term disability

Life insurance

Wellbeing benefits

10 paid holidays, 10 paid sick days,

17 days of Paid Personal Time

Descripción de la vacante

TikTok is seeking a Research Scientist Graduate (Conversational AI) in San Jose for a 2027 start. You will work on agent runtimes, memory, tool orchestration, and safe, scalable conversational systems. The role emphasizes self-evolving loops, evaluation rigor, and production-grade guardrails.

Ideal candidates hold a PhD in CS/EE/Math/Stats, strong Python plus one of C++/Go/Rust/Java, and hands-on ML/NLP with LLMs or agents. Collaboration across teams and data-driven iteration is essential.

Formación

• PhD in CS/EE/Math/Stats or related field.
• Strong Python and one of C++/Go/Rust/Java.
• Solid ML/DL/NLP fundamentals with hands-on LLMs or agents.
• Basic statistical literacy: confidence intervals and p-values.
• Ability to read a paper or engineering blog and turn into code.

Responsabilidades

• Architect agent runtime and loop control; implement safe pathways and guardrails.
• Develop memory and context management for long multi-turn conversations.
• Lead data flywheel: SFT/DPO/RLHF and distillation into smaller models.
• Manage tools, skills, and MCP with governance and on-demand loading.
• Design evaluation strategies: LLm-as-judge, cross-family panels, and canary tests.
• Build auditable pipelines: candidate registry, runtime read-back, change lineage.

Conocimientos

Python

C++/Go/Rust/Java

ML/NLP fundamentals

Statistical literacy

Code-to-paper translation

Educación

PhD in Computer Science, Electrical Engineering, Mathematics or Statistics

Herramientas

vLLM

TensorRT-LLM

KV / prompt caching

Descripción del empleo

TikTok is seeking a Research Scientist Graduate (Conversational AI) in San Jose for a 2027 start. You will work on agent runtimes, memory, tool orchestration, and safe, scalable conversational systems. The role emphasizes self-evolving loops, evaluation rigor, and production-grade guardrails.

Ideal candidates hold a PhD in CS/EE/Math/Stats, strong Python plus one of C++/Go/Rust/Java, and hands-on ML/NLP with LLMs or agents. Collaboration across teams and data-driven iteration is essential.