Senior Data Engineer, Ads

About the role

Discord has a highly engaged community of millions of daily active users who use the platform for many different reasons, but there’s one thing that nearly everyone does: play video games. Discord plays a uniquely important role in the future of gaming, and we are focused on making it easier and more fun for people to hang out before, during, and after playing games. Discord seeks a seasoned Data Engineer, focusing on our advertising product data. In this role, you will drive technical vision and strategy for ads data engineering in support of our ML initiatives while building and maintaining

Requirements

• Develop data quality frameworks, monitoring systems, automated anomaly detection, and SLA infrastructure for critical ads pipelines at massive scale.
• Proactively identify foundational data infrastructure gaps - including those with broad implications across ML, measurement, and reporting - and design scalable, canonical solutions that multiple team
• Build systems from scratch in a rapidly evolving, greenfield advertising data environment - making sound architectural decisions with incomplete information and balancing short-term delivery with long
• Drive alignment across Data Science, ML Engineering, Ads Product, and GTM teams through clear narratives that connect data infrastructure decisions to business outcomes and revenue impact.
• Mentor engineers through technical challenges, code and design reviews, and ownership of complex projects - contributing to the culture and engineering standards of the Data Engineering team team.
• 5+ years of hands-on experience writing production code and architecting data pipelines with high-volume consumer data in advertising technology domains (ad delivery, ranking, targeting, identity, con
• Deep expertise in digital advertising data engineering - specifically in ads delivery, conversion measurement, attribution pipelines, or ML feature data infrastructure. Experience with Conversion Data
• Demonstrated experience building data models in a greenfield or 0-to-1 environment where requirements change frequently, documentation is sparse, and architectural decisions are made with incomplete i