Drimva

Scale AI

Frontier Agents Engineer (Applied AI)

$180K–$225K

San Francisco, CA +2 · Full-time · Posted Jul 31, 2026

Location
San Francisco, CA
New York, New York, United States
San Francisco, California, United States
Experience
4+ years of software engineering, machine learning, or applied AI experience.

About the job

Build and deploy production AI agents and enterprise AI systems, translating frontier research and customer needs into reliable solutions and measuring their business impact.

What you'll do

  • Design, evaluate, and deploy AI agents, retrieval and memory systems, and multi-agent workflows for enterprise use.
  • Run experiments and build evaluation, monitoring, and safety systems to improve reliability, performance, and cost.
  • Work with enterprise customers to translate business challenges into prototypes and scalable production systems.

What they ask for

  • 4+ years of software engineering, machine learning, or applied AI experience.
  • Strong Python programming skills and experience building production AI systems using LLMs.
  • Experience with modern AI tooling, including OpenAI, Claude, MCP, agent frameworks, vector databases, or retrieval systems.
  • Strong understanding of machine learning fundamentals and modern language models; experience evaluating AI systems with quantitative metrics.

Nice to have

  • Experience building production AI agents or autonomous systems and designing evaluation frameworks for LLMs or agentic systems.
  • Experience with RAG, semantic search, knowledge graphs, model optimization, distributed systems, cloud platforms, or enterprise customer engineering.

Pay

$180,000 — $225,000 USD (base salary)

Tools

Python, OpenAI, Claude, MCP, AWS, Azure, GCP, Docker

AI focus

Applied AI: build production agents and multi-agent systems; integrate frontier models with retrieval, memory, structured knowledge, traditional ML, and enterprise software; evaluate models and improve systems using experiments and real-world feedback.

Model-written summary · the original posting is authoritative

First seen Oct 10, 2026 · Last verified Nov 5, 2026 (UTC)