Senior AI Platform Engineer - Agentic Systems

Roblox Roblox · Consumer · San Mateo, CA · Software Engineering

Senior AI Platform Engineer focused on Agentic Systems, responsible for engineering and scaling foundational data infrastructure including Knowledge Graphs and Vector Databases to power AI applications. The role involves designing, developing, and operating these systems at massive scale, building APIs for integration into agentic and RAG applications, and researching advancements in VectorDB and Knowledge Graph technologies. The goal is to create a cutting-edge AI/ML platform for a large-scale consumer product.

What you'd actually do

  1. Design, develop, and operate Knowledge Graph systems for serving, ingesting data, and executing graph queries at massive scale for critical ML applications (e.g., recommendations, fraud detection, safety, and social proof).
  2. Own and significantly improve our Vector Database infrastructure, scaling it to support 100B+ vector searches and ensuring high-performance, low-latency lookups.
  3. Build robust, highly available APIs and services to seamlessly integrate Knowledge Graphs and Vector DBs into new agentic and RAG (Retrieval-Augmented Generation) applications.
  4. Research, evaluate, and pioneer the adoption of the latest advancements in VectorDB technology, Knowledge Graphs, and Ground Truth Data labeling practices to define new industry standards.
  5. Develop comprehensive evaluation frameworks to measure agentic performance, safety, and reliability.

Skills

Required

  • designing, building, and operating large-scale distributed systems
  • production-grade ML infrastructure
  • ML Data Infrastructure
  • Vector Databases
  • Knowledge Graphs
  • APIs and services

Nice to have

  • Ground Truth Data labeling practices

What the JD emphasized

  • massive scale
  • 100B+ vector searches
  • agentic
  • RAG
  • VectorDB
  • Knowledge Graphs
  • AI Platform

Other signals

  • building foundational data infrastructure for AI applications
  • engineering and scaling core AI infrastructure
  • designing, building, and scaling core AI infrastructure from the ground up
  • scaling Vector Database infrastructure to support 100B+ vector searches
  • integrating Knowledge Graphs and Vector DBs into new agentic and RAG applications