Drimva

Snowflake

Staff Software Engineer, Frontier Security Team

$236K–$295K

US-CA-Menlo Park · Staff · Full-time · Posted Sep 18, 2026

Experience
9+ years of software engineering experience, including technical leadership of complex production systems.
Education
Bachelor's degree in Computer Science, Engineering, or a related field, or equivalent practical experience.

About the job

Leads the design and development of an agent execution harness and evaluation platform, taking AI capabilities from prototype to reliable customer-facing services.

What you'll do

  • Architect and build infrastructure for multi-step agent workflows, including tool execution, state, memory, permissions, and recovery.
  • Own agent quality through evaluation systems, datasets, graders, experiments, production analysis, and regression prevention.
  • Set technical direction across teams, lead broad-scope projects, and mentor engineers while contributing to implementation and debugging.

What they ask for

  • 9+ years of software engineering experience, including technical leadership of complex production systems.
  • Direct experience shipping and operating LLM applications, AI agents, or model-backed workflows in production.
  • Strong background in distributed systems and experience building agent, workflow, developer-platform, or evaluation infrastructure.
  • Fluency in Python and proficiency in at least one of Java, Go, Rust, or TypeScript.

Nice to have

  • Evaluation or observability infrastructure for agentic coding, data engineering, or analytics systems.
  • Experience with human evaluation, multi-agent orchestration, simulations, retrieval systems, or sensitive enterprise data.

Pay

$236K – $295K (base salary; currency not stated)

Tools

Python, Java, Go, Rust, TypeScript, Kubernetes, SQL

AI focus

Builds production-grade LLM applications, intelligent agents, agent infrastructure, and evaluation systems; improves agent quality, safety, efficiency, and reliability.

Model-written summary · the original posting is authoritative

First seen Oct 27, 2026 · Last verified Nov 5, 2026 (UTC)