Machine Learning Engineer III, Metadata Extraction

Box is a leader in Intelligent Content Management, enabling organizations to manage content lifecycle, secure critical content, and transform workflows with enterprise AI. Founded in 2005, the company serves global organizations and is headquartered in Redwood City, CA, with offices worldwide. The role is part of the Metadata Extraction team that builds AI-powered capabilities to transform unstructured enterprise content into structured metadata using technologies like OCR and LLM-based extraction agents.

• 3+ years of experience building, deploying, or operating production software, data, or machine learning systems.
• Strong programming skills in Python, writing production-grade, readable, testable, and maintainable code.
• Practical experience with LLM-based agents lifecycle including data preparation, prompting, RAG, evaluation, deployment, and monitoring.
• Experience with LLM frameworks and libraries such as LangChain, LangGraph, PyTorch, Pydantic, Pandas, and NumPy; comfortable calling LLM APIs like Gemini, Claude, OpenAI.
• Understanding of core machine learning concepts and ability to balance model performance with reliability, latency, scalability, and cost.
• Experience building APIs, services, pipelines, or distributed systems in a cloud environment.
• Familiarity with CI/CD, observability, testing, and production support practices.
• Effective communication skills with technical and non-technical partners.
• Ownership mindset with ability to solve ambiguous problems and collaborate across teams.

Nice to have:

• Experience with MLOps tooling, model monitoring, feature stores, workflow orchestration, or experiment tracking.
• Experience with generative AI in production including multi-provider gateways, semantic search, or structured extraction.
• Experience with multimodal intelligent document processing workflows.
• Experience with major cloud platforms such as GCP, AWS, Azure, and technologies like Kubernetes and Terraform.
• Experience with SQL and large-scale analytics (e.g., BigQuery) for quality, cost, and incident investigation.
• Experience in enterprise software, security, search and retrieval, or content intelligence.
• Experience mentoring engineers or contributing to technical direction.

• Design, build, deploy, and maintain production LLM agents, machine learning models, and services.
• Develop robust data preparation pipelines, prompt and tool design, and agentic workflows.
• Write high-quality, maintainable, and well-tested Python code for production systems.
• Partner with Product, Software Engineering, Data, and Infrastructure teams to translate customer and business needs into ML-powered capabilities.
• Evaluate and improve agents' quality using offline and online metrics, eval sets, experiments, and feedback loops.
• Monitor production ML systems for quality, reliability, latency, cost, and data or model drift; investigate and resolve issues.
• Contribute to the architecture of scalable services and pipelines supporting real-time and batch ML workloads.
• Create clear technical designs, documentation, tests, and operational runbooks.
• Share knowledge, contribute to engineering best practices, and improve code quality and operational excellence.
• Participate in an on-call rotation and assist with triaging production issues.

Equal opportunity employer; no discrimination based on race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, disability, or any other protected characteristic under applicable human rights legislation.