Applied AI Engineer

Role overview Build AI features for a consumer application designed to bring useful intelligence into everyday communication and coordination with little need for complex prompting. The role covers the path from LLM and agent design through backend integration, evaluation and production improvement, in close partnership with product and engineering colleagues.

Responsibilities - Develop and ship features using language models, tool calls and agent workflows. - Create prompting, context, memory and orchestration systems for multi-step tasks. - Build backend APIs and integrations between models, the application and external tools. - Evaluate task completion, response quality and reliability, then use results to improve the system. - Investigate model outputs, tool failures and agent behavior through logs and evaluations. - Improve response speed, inference costs and service reliability.

Requirements - At least three years of software or machine learning engineering experience, including AI applications. - Strong Python skills and experience building production backend systems. - Practical experience with LLM APIs, prompt design, tool use and agent workflows. - Understanding of model capabilities, limitations, inference and context management. - Experience evaluating AI outputs and debugging model-driven workflows. - Familiarity with APIs, databases, asynchronous processing, distributed systems and production operations.

Nice to have - Experience with retrieval-augmented generation, embeddings, vector databases or retrieval systems. - Experience with open-source models such as Llama or Qwen, and serving tools such as vLLM. - Model fine-tuning with PyTorch or JAX. - Experience building consumer AI products, especially messaging or productivity applications.