AI Data Engineer

Dscout is a leading UX research platform trusted by top brands across various industries. As an AI Data Engineer, you will design and build data pipelines and work closely with machine learning and analytics teams to ensure the reliability and quality of data that powers AI features.

Responsibilities:

• Design, build, and own the data pipelines that move and transform data from our application and third-party sources into relational databases - keeping them reliable and well-modeled as volume and complexity grow
• Partner with analytics engineering to build the data models and reporting that give researchers and teams real visibility into how their work is performing
• Own data quality as a first-class concern across ingestion, modeling, and reporting. Catch problems before they reach a model, a dashboard, or a user, and fix them
• Build and ship production AI systems the data infrastructure and services that ML features run on
• Design and own evaluation systems that tell us whether an AI feature is ready to ship and holding up over time: eval harnesses, test datasets, and production monitoring built as software, not one-off analyses
• Set the standard for how data work gets done. Write clearly, share context early, and make the people around you faster

Requirements:

• 5+ years of experience in data engineering, with meaningful exposure to ML systems in production
• Deep data engineering experience: you've personally built and owned pipelines that move data from application sources into a warehouse at scale, and you know what breaks, when, and why
• Strong Python skills and fluency across the data stack (we use Snowflake and Postgres)
• Experience with orchestration tools like Airflow, Dagster, or similar
• Experience working with cloud computing environments like GCP or AWS
• A track record of working closely with analytics engineers or data analysts to build reliable, well-documented data models
• Hands-on experience shipping AI or ML systems that real users depended on in production, and owning what happened after launch
• A genuine point of view on evaluation: you treat evals as something you build, not a report you write
• A high-agency mindset. You can take an ambiguous problem and drive it to a working outcome without a fully-scoped ticket
• Fluency across the data lifecycle, from production-facing features to the internal analytics that drive decision-making
• Comfort working across the data lifecycle, from production-facing features to the internal analytics that drive decision-making
• Comfort using AI coding tools (Cursor, Claude Code, Copilot, or similar) as a real part of your workflow
• Experience with dbt or similar data modeling frameworks
• Familiarity with LLM evaluation and observability tooling
• Exposure to MCP-based tooling or agentic data workflows