Software Engineer, Agent Data Platform

Sierra Sierra · AI Frontier · San Francisco, CA · Engineering

Software Engineer, Agent Data Platform role focused on building the data foundations, real-time pipelines, and personalization primitives to make AI agents smarter and improve customer experiences. The role involves architecting and building scalable data systems, including eventing infrastructure, ETL pipelines, data lakehouse, and OLAP querying, as well as developing systems for agent memory, CDP primitives, and optimization loops for personalization.

What you'd actually do

  1. Architect and build our core platform to handle data at scale with low latency. This includes our real-time eventing infrastructure, streaming and batch ETL pipelines, and our Iceberg-based data lakehouse.
  2. Own the systems for interactive OLAP querying, orchestration, and experimentation, ensuring our data is trustworthy, fast, and easy to use for the entire organization.
  3. Develop the systems for long-term agent memory, build reusable Customer Data Platform (CDP) primitives, and implement the optimization loops that allow our agents to personalize interactions and demonstrably lift business outcomes.

Skills

Required

  • design, build, and operate large-scale data systems
  • processing terabytes or petabytes of data
  • Spark, Flink, Trino/Presto, Iceberg or Hudi
  • backend and distributed systems fundamentals
  • Go, Scala, or Python
  • data modeling best practices
  • full-stack engineer (e.g., TypeScript/React and a service layer like Go)
  • 4-7+ years of hands-on development experience
  • building and shipping production systems or products

Nice to have

  • Deep experience with event streaming platforms (e.g., Kafka, Kinesis) and real-time data processing.
  • Experience with modern data visualization libraries (e.g., ECharts, D3.js)
  • Production experience with recommender systems or optimization loops (e.g., multi-armed bandits, ranking).
  • A track record of leading complex technical projects or mentoring other engineers.

What the JD emphasized

  • building the systems making our AI agents measurably smarter with every interaction
  • building the real-time pipelines, analytics products, and deep personalization primitives that turn millions of conversations into business outcomes
  • long-term agent memory
  • reusable Customer Data Platform (CDP) primitives
  • optimization loops that allow our agents to personalize interactions and demonstrably lift business outcomes

Other signals

  • building the systems making our AI agents measurably smarter
  • real-time pipelines, analytics products, and deep personalization primitives
  • long-term agent memory
  • reusable Customer Data Platform (CDP) primitives
  • optimization loops that allow our agents to personalize interactions