The posting, in Starburst's own words
archived Sep 30, 2026About Starburst
Starburst delivers enterprise intelligence at scale by giving organizations secure, governed access to all their data, wherever it lives. Built for distributed data environments, Starburst helps enterprises power AI and analytics without the cost and complexity of traditional data consolidation. With open standards including Trino and Apache Iceberg, Starburst enables trusted access to complete enterprise context while helping organizations avoid vendor lock-in. Leading global enterprises trust Starburst to fuel AI, analytics, and enterprise intelligence. Learn more at starburst.ai .
About the role
We build the AI layer for Starburst's products, including AIDA. We design agents that let users ask questions in natural language and get accurate, grounded answers backed by their actual data. We operate with startup speed inside an enterprise company, shipping weekly and measuring results. This is the first dedicated agent engineering hire on the team. You will raise the bar on agent reliability, turning shipped features into systems users depend on. You will approach AI agents as production software systems, built with the same rigor as any large-scale backend platform. This is not a research role or a prototyping role. You will ship agentic features that work with messy real-world data, undocumented schemas, and diverse access patterns. You have already shipped agentic or LLM-powered systems at an AI startup or AI-focused product team. You bring proven patterns, knowledge of real failure modes, and operational lessons that would take the team months to learn through trial and error. You write code every day and ship every week.
Read the full posting ↓
As an AI Agent Engineer at Starburst, you will:
Design and build agent architectures: tool orchestration, planning, memory, multi-step reasoning, error recovery Ship production-quality agent features with clear ownership from design through deployment Close the gap between demo and production: handle partial schemas, inconsistent metadata, mixed data formats, and unreliable source systems Introduce proven patterns for agent reliability: structured error handling, fallback chains, observability, guardrails Build evaluation suites that catch hallucinations, regressions, and grounding gaps before they reach users Collaborate with the AI Research Engineer to integrate grounding and retrieval into agent workflows Drive technical standards for agent development through code, not documents Turn improvements into measurable gains in task completion rate, accuracy, and time-to-value