SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 210,000 - 260,000 / annual
Reality Defender is a cybersecurity company specializing in deepfake and AI-generated media detection. The company uses a patented multi-model approach to detect fraud, disinformation, and harmful deepfakes in real time across video, audio, imagery, and text. Backed by top-tier investors including DCVC, Y Combinator, Booz Allen Hamilton, IBM, and Accenture, Reality Defender serves leading enterprises, financial institutions, and governments.
As Staff Data Platform Engineer, you will own the end-to-end AI data platform that powers Reality Defender's vision, audio, and multimodal research. You'll architect and scale the company's Databricks lakehouse and distributed processing infrastructure, handling continuous ingestion, preprocessing, governance, and export of image, video, audio, and text data at high volume. Key responsibilities include:
- Own the complete data platform lifecycle: ingestion, preprocessing, governance, and export of multimodal data at scale
- Complete and scale the Databricks lakehouse and distributed processing stack (Spark, Ray, or similar), making strategic build-vs-buy decisions
- Build reusable infrastructure for in-house synthetic data generation, enabling researchers to produce deepfake variants on demand
- Ensure all datasets are reproducible and traceable through versioning, lineage, metadata, quality checks, and stable train/eval splits
- Own pipeline reliability, observability, and cost efficiency as core platform metrics
- Partner with ML researchers to understand data needs and build self-serve tools that accelerate model iteration
- Mentor engineers and establish data engineering best practices across code review, testing, documentation, and runbooks
- Collaborate with Security and compliance teams to ensure data handling, licensing, and access controls meet enterprise and government standards
- Shape the long-term data platform roadmap with engineering and research leadership
Within 12 months, you'll deliver a mature self-serve data platform used by researchers to create and export datasets independently, scale synthetic data generation across vision, audio, and multimodal research, and establish reproducible evaluation datasets that enable confident model comparison.
This is a full-time role reporting to the Director of Data, based at NYC HQ with remote (U.S.) flexibility.
REQUIREMENTS:
- 8+ years building data platforms or data infrastructure, including staff- or lead-level ownership of production pipelines handling high-volume data generation, ingestion, or delivery
- Deep hands-on experience with distributed processing (Spark, Ray, or similar) and lakehouse platforms, ideally Databricks, on cloud object storage (S3, GCS, or Azure Blob)
- Production experience with Databricks, Unity Catalog, Delta Lake, and Spark on AWS, including performance tuning, access controls, and operational troubleshooting
- Experience working with large unstructured media (video, audio, images), including large file handling, decoding/transcoding, and metadata at scale
- Track record building reproducible ML datasets: versioning, lineage, data quality validation, and orchestration (Airflow, Dagster, Databricks Workflows, or similar)
- Strong Python and SQL with judgment to own ambiguous platform problems end-to-end on a lean team
- Infrastructure and deployment automation using Terraform and CI/CD, including AWS storage, IAM, compute, Databricks resources, and distributed processing environments
Nice to have:
- Experience building synthetic data generation pipelines or running generative models (diffusion, voice cloning/TTS, video generation) at scale
- Background supporting ML research teams with training and evaluation data, especially for multimodal models
- Experience in security, fraud, trust & safety, or adversarial domains, or at Series A/B startups where scope outpaces headcount
