Senior AI Software Engineer, Agent Systems

Scale%20army%20careers; United States; Remote; USD 120,000 - 180,000; Full time

Job summary

We are seeking a highly skilled software engineer to design and implement long-running multi-agent loops, with explicit budgets, triggers and verification. You will build swarms, orchestrators and shared state, and own the end-to-end platform infrastructure using TypeScript, Node.js, NestJS and Prisma.

You'll deploy to AWS, manage containers with Docker, implement robust CI/CD pipelines, and ensure observability, tracing and human-in-the-loop reviews for governance and safety in production-grade

Qualifications

• 5+ years of experience writing production-grade software.
• Proven experience in loop engineering (designing agent loops with budgets and self-verification).
• Experience building or operating multi-agent/swarm systems with orchestration and shared state.
• Fluency with modern agent tooling (Claude Code, MCP, tool/function calling, reason-act-observe patterns).
• Strong proficiency in TypeScript and Node.js.
• Experience with service frameworks like NestJS and typed data layers like Prisma.
• Proficiency with AWS (ECS/Fargate), Docker, and CI/CD pipelines.

Responsibilities

• Design self-running agent loops with explicit triggers, scopes, budgets, and stop conditions
• Build multi-agent swarms using orchestrators and specialized agents with clear task ownership
• Implement first-class verification systems, including checks, evals, and reviewer agents
• Manage agent state and persistent memory to ensure long-running tasks survive restarts
• Develop and ship platform infrastructure including APIs, services, and queues using TypeScript and Node
• Deploy and manage services on AWS with robust testing, tracing, and observability
• Maintain human-in-the-loop checkpoints for plan approval and pull request reviews

Skills

TypeScript

Node.js

AWS

CI/CD

React

Tools

NestJS

Prisma

Docker

Job description

Responsibilities

• Design self-running agent loops with explicit triggers, scopes, budgets, and stop conditions
• Build multi-agent swarms using orchestrators and specialized agents with clear task ownership
• Implement first-class verification systems, including checks, evals, and reviewer agents
• Manage agent state and persistent memory to ensure long-running tasks survive restarts
• Develop and ship platform infrastructure including APIs, services, and queues using TypeScript and Node
• Deploy and manage services on AWS with robust testing, tracing, and observability
• Maintain human-in-the-loop checkpoints for plan approval and pull request reviews

Requirements

• 5+ years of experience writing production-grade software
• Proven experience in loop engineering (designing agent loops with budgets and self-verification)
• Experience building or operating multi-agent/swarm systems with orchestration and shared state
• Fluency with modern agent tooling (Claude Code, MCP, tool/function calling, reason-act-observe patterns)
• Strong proficiency in TypeScript and Node.js
• Experience with service frameworks like NestJS and typed data layers like Prisma
• Proficiency with AWS (ECS/Fargate), Docker, and CI/CD pipelines
• Ability to work during U.S. business hours

Preferred Qualifications

• Experience managing token and cost budgets for parallel agents at scale
• Knowledge of distributed systems, queues, and event-driven design
• Experience with React for agent-facing interfaces
• Background in developer tooling or orchestration frameworks
• Awareness of SOC 2 or ISO 27001 standards

About the Company

We build and ship agent platforms that perform real work in production, focusing on coordinated swarms and self-running loops rather than simple prompt engineering.