AI Agent Algorithm Engineer Graduate (Global E-Commerce, Conversational AI) - 2027 Start (PhD)

TikTok
Singapore
Workplace: OnsiteFull timeFunction: Research & Scientific (R&D)Education: phdSkills: []

Build and evolve an agent runtime for TikTok’s global e-commerce customer service, powering self-evolving agents across 30+ languages. You’ll engineer agentic memory and context (retrieval-then-load, compaction, masking), post-train and fine-tune LLMs, design tool/MCP integrations, and create LLM-as-judge evaluation harnesses that close offline-to-online gaps. Own end-to-end delivery measured on real CSAT and containment metrics.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
TikTok
TikTok
1 day ago

AI Agent Algorithm Engineer Graduate (Global E-Commerce, Conversational AI) - 2027 Start (PhD)

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 13 hours agoStatus: Live

Job Summary

Build and evolve an agent runtime for TikTok’s global e-commerce customer service, powering self-evolving agents across 30+ languages. You’ll engineer agentic memory and context (retrieval-then-load, compaction, masking), post-train and fine-tune LLMs, design tool/MCP integrations, and create LLM-as-judge evaluation harnesses that close offline-to-online gaps. Own end-to-end delivery measured on real CSAT and containment metrics.
Location: Singapore
Workplace: Onsite
Employment Type: Full time
Job Function: Research & Scientific (R&D)
Seniority: Graduate level

Key Responsibilities

  • •Build and evolve the agent runtime (harness/agent loop) including skill/tool orchestration, loop control, progressive disclosure, and behavior-level guardrails.
  • •Engineer agentic memory and context for long multi-turn conversations, including context compaction/summarization and retrieve-then-load retrieval.
  • •Post-train and fine-tune LLMs (SFT/DPO/RL) and build the data flywheel that converts served conversations into training/eval/retrieval signals.
  • •Design and integrate tools/Skills and MCP connectors (tools-as-APIs), including skill/tool search for large tool inventories.
  • •Create evaluation systems using LLM-as-judge with human-agreement calibration, and build self-evolving loops via replay, user simulation, and automated improvement mechanisms.

Key Requirements

  • •Completing or having recently completed a PhD in CS/AI/Math or another quantitative field.
  • •Strong Python plus one of C++ / Go / Rust / Java.
  • •Solid ML/DL/NLP fundamentals with hands-on experience with LLMs or agents (research, coursework, internship, open source, or serious side project).
  • •Ability to read papers or engineering blogs and turn them into working code.
  • •Commit to an onboarding date by the end of the year and state availability/graduation date in the resume.
Education:PhD / Doctorate in CS / AI / Math or quantitative field
Tech Stack:PythonC++GoRustJavaLLMsMLDLNLPSFTDPORLRLHFRLAIFRLVRReward modelingAgent systemsHarnessContext engineeringMCP

Company Brief

TikTok
Short-form video platform that lets users create, share, and discover entertainment content through algorithmic recommendations. It also offers advertising and creator tools for brands, influencers, and businesses.
Industry: Digital Media
Company Size: Enterprise (1,001+ employees)
Growth: Scaleup
Headquarters: Singapore, Singapore
Founded: 2016
WebsiteLinkedIn