Staff ML Research Engineer, Multimodal Structure & Marengo
Salary not listed • Seoul, South Korea • Full-time
Posted 1mo ago
Job Location
Seoul, South Korea
Tech Stack
Remote Work Policy
On-site
Employment Type
Full-time
Categories
AI Research Engineer
About the job
TwelveLabs is building the intelligence layer for video data, enabling machines to understand video content across sight, sound, and motion. Our multimodal AI models power production-scale AI workloads in media, entertainment, sports, security, and government. We are a global company headquartered in San Francisco with offices in Seoul, New York, and London, seeking individuals driven by hard problems who want their work to matter. This role focuses on developing core capabilities for multimodal understanding, specifically structure and semantics, contributing to our unified agentic system, Jockey, and our embedding model, Marengo.
Responsibilities
- Set the technical direction for multimodal structure, including asset organization and relationships.
- Define the architecture, training, and data strategy for next-generation multimodal embedding models.
- Own end-to-end model development from research planning through large-scale distributed training to production evaluation.
- Architect and optimize large-scale training infrastructure, including distributed training pipelines, data processing systems, experiment workflows, and GPU utilization.
- Own production model APIs end to end, from model packaging and API design to inference optimization and reliable operation at scale.
- Drive data strategy by building large-scale curation, filtering, and quality systems for structure and embeddings.
- Define evaluation methods and quality standards for structure, embeddings, and end-to-end multimodal understanding.
Requirements
- Set technical direction for multimodal structure and embeddings.
- Define architecture, training, and data strategy for next-generation multimodal embedding models.
- Own end-to-end model development lifecycle.
- Architect and optimize large-scale training infrastructure.
- Own production model APIs end to end.
- Drive data strategy for curation, filtering, and quality systems.
- Define evaluation methods and quality standards.
- Thrive in ambiguity and identify high-impact research problems.
- Define technical approaches and drive cross-team execution.
- Experience in video segmentation and tracking, temporal and hierarchical modeling, contrastive learning, and multimodal representation learning.
- Experience with large-scale distributed training and robust evaluation frameworks.
- Experience with world-class compute infrastructure, including NVIDIA B300 GPUs.