MLOps / ML Deployment Engineer

About the Role

We are seeking an autonomous MLOps / ML Deployment Engineer to help us bridge the gap between Large Language Models (LLMs) and our internal systems. In this role, you won't just be deploying static models; you will be building the infrastructure that allows LLMs to interact with the real world. Your primary focus will be designing and deploying autonomous AI agents and custom Model Context Protocol (MCP) servers. If you have a strong background in productionizing ML pipelines, building robust microservices, and managing secure API integrations, you already have the foundational skills to succeed here—we will trust you to map those skills to the emerging world of agentic workflows.

What You'll Do

• Productionize LLM Workflows: Translate prototype prompt-engineering and LLM concepts into robust, scalable, production-grade agentic pipelines.
• Build Integration Gateways (MCP): Design and deploy secure, lightweight microservices (using Anthropic's open-source MCP standard) to expose databases, APIs, and internal tools as context for LLMs.
• Orchestration & State Management: Implement reliable workflow orchestration to handle agent reasoning steps, tool calling, and graceful error handling.
• CI/CD & Monitoring: Set up logging, evaluation, and monitoring for deployed LLM interactions to ensure system reliability and data privacy.

What We're Looking For

• ML Deployment & MLOps: Proven track record of deploying, scaling, and monitoring machine learning models or complex backend systems in production environments.
• Backend Mastery & API Design: Expert-level Python (including async programming, FastAPI, or similar) and a strong grasp of REST/gRPC API design.
• Data Integration: Deep experience connecting applications to production data layers, including SQL/NoSQL databases, internal web services, and vector stores.
• Infrastructure & Containerization: Proficient with Docker, modern CI/CD pipelines, and cloud infrastructure (AWS, GCP, or Azure).
• Autonomy & Grit: Experience operating independently as a contractor, with the ability to read documentation for emerging technologies (like MCP) and rapidly implement them.

Nice to Haves

• Familiarity with LLM orchestration tools (e.g., LangChain, LlamaIndex, CrewAI) or traditional workflow tools (e.g., Airflow, Prefect).
• Conceptual understanding of the Model Context Protocol (MCP) or tool-calling architectures.
• Experience with serverless deployments or message brokers (Kafka, RabbitMQ).