Lead Genai Forward Deployed Engineer, Youtube

Google Google · Big Tech · San Bruno, CA +3

Google is seeking a Lead GenAI Forward Deployed Engineer for their YouTube AI Accelerator team. This role focuses on driving AI transformation for YouTube's business operations by redesigning workflows and deploying applied AI solutions. The engineer will build GenAI PoCs, lead the delivery of complex AI applications from prototype to production, and act as a trusted engineering partner. Responsibilities include authoring technical designs, writing code, building front-ends, and developing high-performance eval pipelines and observability frameworks to address AI bottlenecks.

What you'd actually do

  1. Build GenAI PoCs (LLMs, RAG, agentic frameworks) to demonstrate feasibility to stakeholders, translating ambiguous business problems into software solutions.
  2. Serve as engineering Lead delivering complex AI applications, transitioning rapid prototypes to production-grade systems (e.g., multi-agent systems, MCP servers) that drive Return on Investment (ROI).
  3. Act as a trusted engineering partner to product managements and stakeholders, co-creating tool roadmaps that enable YouTube's business operations.
  4. Author technical designs, write clean/maintainable code, build front-ends, and deliver from scoping to deployment.
  5. Build high-performance eval pipelines and observability frameworks to resolve AI bottlenecks (data readiness, system integration, state-management) while ensuring accuracy, safety, compliance, and latency.

Skills

Required

  • 9 years of experience building APIs or web applications
  • designing scalable microservice or system architectures
  • backend data pipelines
  • distributed systems
  • data privacy
  • full-stack development
  • Python
  • Go
  • Java
  • C++
  • TypeScript
  • building enterprise-grade applied AI solutions
  • working with teams and business stakeholders to create product roadmaps
  • Tech Lead or Engineering Manager experience

Nice to have

  • SRE
  • InfoSec
  • DevOps
  • building LLM evaluation (evals) pipelines
  • observability
  • optimizing LLM-native metrics at scale
  • vulnerability testing
  • security auditing
  • implementing AI safety guardrails
  • enterprise accuracy
  • enterprise AI data infrastructure
  • vector databases
  • embedding generation
  • search architectures
  • data readiness
  • state-management
  • multi-agent systems (ReAct, tool-calling)
  • MCP
  • orchestration frameworks
  • GTM
  • sales workflows
  • CRMs
  • technical adoption of zero-to-one AI products

What the JD emphasized

  • enterprise-grade applied AI solutions
  • AI safety guardrails
  • enterprise accuracy
  • AI bottlenecks

Other signals

  • GenAI transformation
  • applied AI
  • enterprise-grade AI powered solutions
  • AI Accelerator team