About this role
Engineers who thrive on hands-on systems work and take full ownership of delivery will find an ideal home in this position. Abaka AI builds the foundational data pipelines and datasets relied on by over a thousand industry leaders across generative, embodied, and automotive artificial intelligence.
This engineer owns multimodal data projects from initial scoping through final handoff, designing scalable processing systems and establishing rigorous quality standards. The existing data engineering team needs someone capable of turning ambiguous client needs into clear technical plans while scaling infrastructure and operating standards.
What you will do
- Assess incoming data requirements for technical feasibility, risks, costs, and delivery timelines before work begins.
- Define execution plans, resource needs, quality specifications, and written acceptance criteria for assigned projects.
- Coordinate work across engineers, annotation teams, and external vendors by breaking requirements into prioritized tasks.
- Build and improve scalable pipelines for multimodal data sourcing, processing, cleaning, annotation, storage, and delivery.
- Establish measurable quality gates at ingestion and prior to delivery to prevent non-compliant data from reaching clients.
- Operate and optimize the underlying infrastructure, including object storage, compute and GPU resources, and orchestration environments.
- Partner with client-facing and foundation model teams to clarify requirements, communicate tradeoffs, and flag risks early.
What you need
- Three or more years of professional experience in data engineering or building large-scale data systems with production workflow ownership.
- Demonstrated history of delivering datasets or data systems end to end, managing scope, timeline, quality, and cost.
- Hands-on practical experience working with at least two distinct data modalities such as text, images, audio, video, or 3D point clouds.
- Proficiency in designing and operating large-scale data pipelines using distributed processing, cloud object storage, and orchestration tools.
- Working knowledge of pipeline infrastructure, including cloud permissions, storage layout, compute provisioning, and cost monitoring.
- Experience implementing data quality and acceptance processes like sampling plans, measurable gates, or model-assisted quality control.
- Familiarity with data privacy and security practices including access control, anonymization, and provenance tracking.
Nice to have
- Experience preparing pretraining, supervised fine-tuning, RLHF, or evaluation datasets for LLMs or multimodal foundation models.
- Experience with data for autonomous driving, robotics, embodied AI, or other sensor-based domains.
- Experience with large-scale deduplication, data quality scoring, or dataset composition design.
- Experience building annotation platforms or internal data tools, or working with external vendors.
- Experience modeling and optimizing the cost of GPU- or compute-intensive data pipelines.
Benefits
- Equity
- Health coverage
- Dental coverage
- Vision coverage
- Paid time off
- Flexible work schedule
Skills and tech stack
Questions candidates ask
What is the salary range for the Multimodal Data Engineer position?
The base salary range for this position is $120,000 to $225,000 USD annually. Compensation may vary outside this range based on a candidate's qualifications, skills, competencies, and experience.
What kind of benefits are included?
This role is eligible for equity and a comprehensive benefits package, including health, dental, and vision coverage, PTO, and a flexible work schedule.
Where is the team located?
Abaka AI has its headquarters in Silicon Valley with teams in Paris, Singapore, and Tokyo, and this specific role is located in the United States.