AI/ML Intern - Agentic Retrieval & Enterprise RAG

You will join the Generative AI team, focused on building intelligent, production-grade AI features used by enterprise customers. The team values curiosity, clear communication, and strong collaboration. You will work closely with engineers and product partners and contribute insights that help shape technical decisions and system design.

Zoom is building the next generation of enterprise AI systems that power intelligent search, retrieval, and agentic workflows across Zoom products. As an AI Engineer / Machine Learning Engineer Intern, you will work on agentic retrieval and Enterprise RAG systems that operate at production scale and directly impact real users. This role sits at the intersection of large language models, retrieval systems, and production engineering. You will help design, implement, and optimize agentic pipelines that retrieve, reason over, and act on enterprise knowledge with high accuracy, low latency, and strong reliability guarantees.

You will contribute to one or more of the following areas:

Agentic Retrieval and RAG Systems

• Design and implement agentic retrieval pipelines that orchestrate multi-step reasoning, retrieval, and tool or function calls

• Improve end-to-end performance of Enterprise RAG systems, including latency, throughput, and cost efficiency

• Integrate agentic workflows with unified retrieval layers and enterprise data sources

Retrieval Quality and Efficiency

• Optimize retrieval strategies to reduce redundant calls and unnecessary context expansion

• Implement content compression or filtering techniques to improve reasoning efficiency while preserving answer quality

• Analyze retrieval and function-call traces to identify inefficiencies and failure patterns

Production-Grade Engineering

• Build scalable, maintainable ML and retrieval components suitable for production environments

• Add instrumentation, logging, and evaluation hooks to measure system behavior and product impact

• Collaborate with engineers and researchers to transition prototypes into reliable services