Staff Data Platform Engineer

present of AI; New York (NY); On-site; USD 150,000 - 210,000; Full time

Job summary

Reality Defender in New York is seeking a senior data platform engineer who will own the data platform powering AI research, including ingestion, processing, governance and export of image, video, audio, and text assets. You will scale pipelines at scale with Databricks, Spark, and AWS, and translate researcher needs into self-serve data tools.

Experience with large media files and governance decisions is highly valued.

Qualifications

• Long experience running large data pipelines.
• Deep familiarity with Databricks, Spark and AWS.
• Experience handling big media files and governance considerations.

Responsibilities

• Own the data platform feeding AI research from ingestion through processing, governance and export.
• Turn researcher needs into self-serve data tools for experiments and models.
• Build and operate scalable data pipelines for image, video, audio and text assets.
• Collaborate with ML researchers to understand data requirements and impact on model quality.

Skills

Databricks

Spark

AWS

Data pipelines

Tools

Terraform

Job description

This role owns the data platform that feeds the company's AI research, covering image, video, audio and text from ingestion through processing, governance and export. A strong candidate has long experience running large data pipelines, is deeply comfortable with Databricks, Spark and AWS, and has handled big media files. They also bring good judgment on build versus buy choices and enjoy turning researcher needs into self-serve tools.

How it fits the AI era

AI changes this work because the data now includes synthetic media made by generative models, and the platform must produce new deepfake variants on demand as new tools appear. Automation can speed up pipelines and quality checks, but people still decide what data is trustworthy, how it is licensed and handled, and which tradeoffs hold up as volume grows. Ownership of reproducibility, so any model result traces back to exact data, remains a human responsibility.

How it builds AI experience

Someone in this role would gain hands-on experience building the data side of a deepfake detection system, including running generative models for voice, image and video at scale. Working closely with ML researchers shows how training and evaluation data shape model quality. To build toward it, practice making versioned, traceable datasets and try small synthetic data pipelines with open diffusion or text-to-speech models.

Worth showing

• Databricks and Spark at scale
• Large video and audio processing
• Terraform and AWS automation

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Reality Defender

A New York company whose API and SDKs, per the company, scan live calls, meetings, images and video in real time to flag AI-generated deepfakes before they reach a bank, call centre or government system.