Lgairesearch
AI Data Engineer
Gangseo-gu, Seoul, South Korea. Posted 44 months ago
The role
This person will design and develop Agent-based Self-Improvement AI Pipelines connecting model evaluation, data generation, and training experiments. They will also analyze Foundation Model performance and improve the AI Data Platform.
What you will do
- Design and develop Agent-based Self-Improvement AI Pipeline connecting model evaluation, data generation, and training experiments
- Analyze Foundation Model performance and failure cases by benchmark to derive improvement tasks
- Generate and refine synthetic and reference data for training and evaluation and automate Data Quality Control
- Compare and verify performance changes based on data composition and training methods in SFT, RLHF, RLVR, Mid-training, and Post-training processes
- Operate AI Data Platform and advance features such as On-Prem Data Storage, Data Versioning, and Experiment Tracking
Required qualifications
- Understanding of Machine Learning and Deep Learning and experience in AI model and data research and development
- Understanding of LLM-based Agent System structure and Agent Orchestration
- Ability to define model and data problems and connect performance improvement through Experimental Design and Statistical Analysis
- Experience collaborating with researchers and engineers to specify technical tasks and apply them to actual research or service environments through implementation and verification
- Ability to actively use Code Agents and verify output quality to increase development productivity
Nice to have
- Experience publishing first-author papers in top-tier NLP, ML, AI conferences such as ACL, EMNLP, NeurIPS, ICML, ICLR, AAAI
- Experience improving actual model performance through large-scale Training Dataset construction and Data Curation
- Experience directly developing Benchmarks, Automated Evaluators, or Reward Models/Verifiers
- Experience deploying and operating AI services or Data Platforms in production environments to improve safety, processing speed, and cost
Before you apply
- LG AI Research is not a military service designated company, so recruitment and transfer of technical research personnel is not possible
- Document screening, 1st technical interview (online), LG Way Fit Test & coding test, 2nd technical interview (online), 3rd technical interview (online), and final interview (on-site)
About the company
LG AI Research leads AI transformation across industries through research and development of large-scale AI models including EXAONE.