Multimodal Data Engineer — Scalable AI Datasets and Equity

Abaka AI; Mountain View (CA); On-site; USD 120,000 - 225,000; Full time

Benefits offered by this job

Equity

Health benefits

Dental coverage

Vision coverage

PTO

Flexible schedule

Job summary

Abaka AI is seeking a Multimodal Data Engineer in the United States to build systems turning complex data requirements into high-quality datasets for AI teams. You will own projects from requirements through delivery, spanning data sourcing, processing, annotation, quality control, storage, and delivery.

The role involves working with data engineers, annotation teams, and external vendors to ensure datasets meet specifications. Strong ownership and hands-on systems work are essential.

Qualifications

• 3+ years of data engineering or large-scale data systems experience with production data workflows.

Responsibilities

• Assess data requirements for feasibility and define acceptance criteria before work begins.
• Own end-to-end delivery for projects from scoping to handoff.
• Build scalable pipelines for multimodal data sourcing, processing, and delivery.
• Establish quality gates and prevent non-conforming data from delivery.
• Track project status with dashboards and dataset indexes.
• Support infrastructure behind data workflows including storage and compute.
• Automate recurring workflows and build internal tooling.
• Collaborate with client-facing and foundation model teams to clarify data needs.

Skills

Data engineering

Large-scale pipelines

Multimodal data

Spark/Ray/Flink

Airflow/Dagster/Argo

Cost optimization

Data quality

Security & privacy

Tools

Spark

Ray

Flink

Airflow

Dagster

Argo

云对象存储

Job description

Abaka AI is seeking a Multimodal Data Engineer in the United States to build systems turning complex data requirements into high-quality datasets for AI teams. You will own projects from requirements through delivery, spanning data sourcing, processing, annotation, quality control, storage, and delivery.

The role involves working with data engineers, annotation teams, and external vendors to ensure datasets meet specifications. Strong ownership and hands-on systems work are essential.