AI Engineering Intern

About this role

About the role: - Build and refine AI agent systems that orchestrate multiple tools and services to solve complex problems - Work with large language models to design and test prompts, function calling mechanisms, and retrieval-augmented generation pipelines - Develop and maintain agent frameworks, focusing on execution loops, state handling, and error recovery - Integrate the Model Context Protocol (MCP) to extend agent capabilities through custom servers and tools - Debug and trace LLM workflows to understand model behavior and optimize performance What you'll work on: - Creating evaluation frameworks and metrics to measure the correctness and quality of agent outputs - Building REST APIs and asynchronous Python services that power agent orchestration - Implementing tool calling and middleware layers that connect agents to external systems - Working with structured data formats like JSON and leveraging Pydantic for type safety - Testing your code through unit tests and version control with Git What we're looking for: - Solid grasp of Python, object-oriented programming, and asynchronous programming patterns - Hands-on experience with at least one agent framework such as LangGraph or DeepAgents, including tool integration and agent orchestration - Knowledge of how agent execution loops work, including state management and error handling - Understanding of MCP architecture with practical experience building or connecting MCP servers - Ability to use tracing and monitoring tools to instrument and debug LLM and agent systems - Familiarity with REST API design and experience working with JSON - Basic knowledge of databases and comfort with Git workflows and debugging Nice to have: - Experience with Azure OpenAI, OpenAI APIs, Anthropic, or open-source language models - Exposure to monitoring platforms like MLflow, LangSmith, Langfuse, or OpenTelemetry - Familiarity with FastAPI, Pydantic, PostgreSQL, MongoDB, or vector databases - Knowledge of Docker, cloud deployment, and CI/CD pipeline setup - Understanding of multi-agent systems, human-in-the-loop workflows, or persistent agent memory - Background in document intelligence, information extraction, or enterprise workflow automation