Design and build a production multi-agent LLM framework and LLM-powered pipelines (text-to-SQL, NLU, structured reasoning). Implement RAG, prompt systems, evaluation and CI/CD for AI features, instrument and optimize model performance, and collaborate with product to deliver user-facing capabilities.

Work on production features for an AI-native fleet management platform: write Python/JavaScript/SQL code, design data pipelines and analytics, and gain hands-on experience with web development and AI/ML workflows.