About the job

Responsibilities:

• Build and maintain LLM based pipelines for processing customer requests
• Design and implement data ingestion and preprocessing pipelines (emails, PDFs, images, etc.)
• Develop classification and routing logic for customer inquiries
• Apply prompt engineering and continuously improve model outputs
• Implement structured outputs and human-in-the-loop (HITL) workflows
• Ensure production readiness (testing, CI/CD, containerization)
• Contribute to AI system architecture and scalability
• Monitor and improve model performance using LLM observability tools
• Collaborate across teams to integrate AI solutions into existing workflows

Requirements:

• Strong experience with Python (production-grade)
• Hands-on with FastAPI and async programming patterns
• Experience building LLM pipelines (OpenAI SDK, LangChain, LangGraph, or similar)
• Solid prompt engineering experience
• Familiarity with structured outputs and LLM evaluation approaches
• Experience with Docker, CI/CD, and testing practices
• Background in data engineering (pipelines, preprocessing, working with unstructured data)
• Experience with PostgreSQL and REST/event driven APIs
• Hands-on use of AI coding tools (GitHub Copilot, Codex, Claude Code, or similar)
• Nice to have: Experience with scalable AI platforms, Azure services, event-driven architectures, and Infrastructure as Code.