Senior Software Engineer, Agent Infrastructure

Cohere builds large language models and NLP infrastructure specifically designed for enterprise use. Unlike consumer-facing AI products, Cohere focuses on text understanding, generation, and semantic search for business applications including search, classification, summarisation, and retrieval-augmented generation. The company offers both cloud API access and on-premises deployment, which matters to enterprises with data residency requirements.

Engineering at Cohere spans model training and fine-tuning infrastructure, inference optimisation, the API platform, and enterprise deployment tooling. The work is highly technical and sits at the frontier of applied LLM research. Python, CUDA, and PyTorch are central to the ML stack, while the platform and API infrastructure use Go and TypeScript. The company is growing rapidly and engineers have significant ownership over technical direction.