DoorDash USA
Software Engineer, Data and AI Platform
About this role
DoorDash seeks a Software Engineer to build AI-powered data analytics products used across the company. You'll own the full stack from user questions to validated answers, combining semantic layers, agentic features, and reporting infrastructure across Snowflake, Spark, and Clickhouse.
What you'll do
- Design and ship agentic analytics features for thousands of users with emphasis on accuracy and data governance
- Build BI authoring, dashboard runtime, drill-down, filtering, scheduling, and alerting systems
- Develop AI-assisted query authoring and generated charts with governance controls
- Own query planning, caching, federation across multiple engines, and safe execution under bursty agent traffic
- Evolve semantic layer with metric/dimension modeling, lineage, and metadata platforms for discovery
- Create self-serve tooling for building, evaluating, and improving AI assistants
What they're looking for
- Backend development in Go and Python
- LLM-powered feature shipping with retrieval, grounding, and eval systems
- Data platform experience (semantic layers, discovery, reporting, caching)
- AWS and Kubernetes in production
- MCP, agent frameworks, or tool-calling architectures
- Analytics/BI tooling and multi-tenant data governance
- Prompt optimization and context management at scale
- Full-stack development (preferred)
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
DoorDash USA
DoorDash USA is building autonomous delivery systems including drones and robots, along with internal infrastructure platforms to support large-scale operations. The company is hiring robotics engineers, autonomous systems specialists, infrastructure engineers, and platform software engineers to develop flight control systems, mapping and localization capabilities, and distributed computing platforms.
View all jobs at DoorDash USALikely interview questions
- Walk us through an LLM-powered feature you shipped to production—how did you handle retrieval grounding and evaluation?
- Describe your experience building or extending semantic layers; how did you handle metric definition governance and lineage?