Software Engineer, Agents & Automations

Cohere Cohere · AI Frontier · London, United Kingdom · Product

Software Engineer role focused on building the core platform for Cohere's North AI workspace, specifically the Agents & Automations features. This involves developing a workflow builder, execution engine, integrations, debugging tools, observability, and evaluation systems to enable enterprises to create and manage AI-powered workflows and agents that interact with workplace tools. The role requires end-to-end ownership of features, working across the full stack, and making practical technical decisions to ship reliable software that augments or automates business processes.

What you'd actually do

  1. Design, build, ship, and maintain core capabilities for North’s Agents & Automations platform.
  2. Build product and platform features that help users create, run, debug, evaluate, and improve agents and automations.
  3. Own features end-to-end, from technical design through implementation, testing, launch, and iteration.
  4. Work across the stack, from frontend product surfaces to backend systems, depending on what the product needs.
  5. Make practical technical decisions that balance speed, quality, depth, and user impact.

Skills

Required

  • Python
  • TypeScript
  • React
  • Postgres
  • Kubernetes
  • backend development
  • frontend development
  • product development
  • data development
  • infrastructure development
  • debugging
  • testing
  • collaboration
  • communication

Nice to have

  • experience with every part of the stack

What the JD emphasized

  • built, shipped, and operated production software used by real users or customers
  • strong fundamentals
  • product judgment
  • agency
  • learning ability

Other signals

  • building AI systems
  • deploying frontier models
  • AI workspace platform
  • AI agents with workplace tools
  • workflow builder
  • execution engine
  • integrations
  • debugging tools
  • observability
  • evaluation systems
  • feedback loops
  • augment or automate real business workflows
  • natural-language intent into a working automation
  • richer analytics for deployed agents
  • debugging tools for failed runs
  • workflow primitives