Software Engineer, Agents

SimSan Francisco, CA

About The Position

You'll own end-to-end development of our core agentic workflow engine and Copilot — the AI assistant that helps developers build and debug workflows using natural language, turning high-level intent into executable, reliable agents. This spans our backend (Next.js), orchestration layer, and all integrations with LLMs and external APIs. Copilot is currently the best-in-class tool for building workflows with natural language. Your job is to keep it there and push it further — making it faster, more reliable, and capable of handling increasingly complex agentic architectures. This is a foundational role. You'll set architecture decisions, reliability standards, and coding patterns that define how the core product works. On a team of 5, your work ships to tens of thousands of developers on fast release cycles.

Requirements

  • Strong generalist engineer who has shipped complex distributed systems or developer tools to production
  • Deep experience with TypeScript/JavaScript and Bun; comfort working in a Next.js monorepo
  • Track record of owning and operating production systems — you've been paged, you've debugged the weird ones, you've made things reliable
  • Experience integrating with LLMs or building on top of foundation model APIs
  • High agency and ownership — on a team this small, you define the roadmap as much as you execute it
  • Strong opinions on architecture, loosely held — you care about building things right but move fast

Nice To Haves

  • Experience with AWS infrastructure is a plus
  • Contributions to open-source projects are a plus

Responsibilities

  • Own the agentic workflow engine: the runtime that executes multi-step, tool-using agent workflows in production
  • Build and evolve Copilot — our natural language interface for creating, editing, and debugging workflows
  • Design and maintain integrations with LLM providers (OpenAI, Anthropic, Google, local models via Ollama) and external APIs
  • Architect the orchestration layer that turns visual flows into reliable, observable agent executions
  • Set technical standards for the codebase — reliability, testing, code quality, and architecture patterns
  • Work across the stack in our Next.js monorepo, shipping to production daily
  • Debug hard problems at the intersection of LLMs, distributed systems, and developer tooling

Benefits

  • Competitive salary + meaningful equity
  • Early team, significant ownership
  • Full benefits
  • Comprehensive health insurance
  • Relocation
  • Visa sponsorship available
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service