About the role
Applied AI Engineer on the Agents & Automations team at Cohere, the enterprise AI company behind its own frontier foundation models. The work centers on North, Cohere's secure AI workspace platform, and on making its agents genuinely useful inside real enterprise workflows.
You would turn raw model capability into reliable, user-facing systems: agent builders, evaluation and observability frameworks, proactive assistants, and new interfaces for human-AI collaboration. A lot of the job is reliability engineering for AI: improving behavior through prompting, context construction, tool-use strategies and feedback loops, then turning production failures and real-task data into stronger eval suites and regression tests. You would work across product, modelling and design.
Highlights
- Frontier applied-AI work at one of the few independent model builders
- Builds agent builders, evaluation and observability frameworks, and proactive assistants on North, Cohere's secure AI workspace platform
- Reliability engineering for AI: prompting, context construction, tool-use strategies, eval suites and regression tests
- Global (Toronto, London, Paris, New York, San Francisco, Berlin), remote-flexible; suits someone who has shipped production LLM products
This recap is dataskew's editorial summary, not the company's copy.