Google AI Architect

Deloitte Deloitte · Consulting

This role focuses on architecting and delivering enterprise AI platforms and applications, specifically leveraging Google Cloud services like Vertex AI and Gemini. The responsibilities include designing, fine-tuning, evaluating, and governing LLM solutions, building RAG and agentic systems, defining end-to-end architectures for the model lifecycle, and implementing security and governance for AI/ML systems. The role also involves cloud-native development and applying design patterns.

What you'd actually do

  1. Architect and deliver enterprise AI platforms and applications on Google Cloud using Vertex AI and Gemini; optimize for scalability, reliability, security, and cost.
  2. Design, fine-tune, evaluate, and govern LLM solutions with Gemini on Vertex AI (prompt/tool/function calling, safety policies, Vector Search, evaluation); implement deployment, inference optimization, and monitoring.
  3. Build RAG and agentic solutions using Vertex AI Vector Search and BigQuery vector; implement context management, retrieval strategies, and observability.
  4. Define end-to-end architectures across data pipelines, feature engineering, model lifecycle, APIs/microservices, and CI/CD/MLOps/LLMOps with Vertex AI Pipelines and Cloud Build.
  5. Lead cloud-native development on GKE, Cloud Run, Pub/Sub, BigQuery, Cloud SQL/Spanner, Memorystore, and Terraform; enforce application and agentic design patterns.

Skills

Required

  • Google Cloud
  • Vertex AI
  • Gemini
  • LLM fine-tuning
  • LLM evaluation
  • LLM governance
  • prompt engineering
  • tool calling
  • function calling
  • safety policies
  • Vector Search
  • deployment optimization
  • inference optimization
  • monitoring
  • RAG
  • BigQuery vector
  • context management
  • retrieval strategies
  • observability
  • data pipelines
  • feature engineering
  • model lifecycle management
  • APIs/microservices
  • CI/CD
  • MLOps
  • LLMOps
  • Vertex AI Pipelines
  • Cloud Build
  • GKE
  • Cloud Run
  • Pub/Sub
  • BigQuery
  • Cloud SQL/Spanner
  • Memorystore
  • Terraform
  • application design patterns
  • agentic design patterns
  • data privacy
  • model poisoning
  • adversarial attacks
  • Gemini safety features
  • enterprise guardrails
  • Cloud Native development
  • Docker
  • Kubernetes
  • serverless functions
  • managed databases
  • GenAI tools

Nice to have

  • Computer Science degree
  • Engineering degree
  • 6+ years experience

What the JD emphasized

  • enterprise AI platforms
  • LLM solutions
  • RAG and agentic solutions
  • end-to-end architectures
  • security and governance for AI/ML systems
  • enterprise-grade AI applications
  • production-level performance and reliability
  • enterprise architects
  • Cloud Native principles
  • GenAI tools
  • security of all AI/ML systems

Other signals

  • Architect and deliver enterprise AI platforms and applications on Google Cloud using Vertex AI and Gemini
  • Design, fine-tune, evaluate, and govern LLM solutions with Gemini on Vertex AI
  • Build RAG and agentic solutions using Vertex AI Vector Search and BigQuery vector
  • Define end-to-end architectures across data pipelines, feature engineering, model lifecycle, APIs/microservices, and CI/CD/MLOps/LLMOps with Vertex AI Pipelines and Cloud Build
  • Implement security and governance for AI/ML systems