Skip to content
LIVE Agentic AI Workshop · 7 NovView workshop →
NETSETOS
My LearningJoin workshop

Explore Netsetos

My Learning
Join workshopAll courses →
Contact Support · info@netsetos.com
top of page

GenAI with GCP

GenAI with GCP cover: a notebook feeding a cloud project with a GPU server, a data table with predictions and a secured endpoint

Group Discussion

This program is connected to a group. You’ll be added once you join the program.

Netsetos Group

Netsetos Group

Private454 Members

Build GenAI systems the way product teams actually ship them — Gemini, Vertex AI, RAG pipelines, and the production patterns you won't find in a hello-world notebook. Included in the Starter plan (₹200). Engineer-taught, India-context aware.

Course syllabus

  1. MODULE 1: GCP Foundation & First LLM Call
    1. 1.1 Setting Up Your GCP AI Project - Create project
    2. 1.2 IAM & Security
    3. 1.3 First Gemini Call - Gemini 3.6 Flash
    4. Q&A - 1.1
    5. Q&A - 1.2
    6. Q&A - 1.3
  2. MODULE 2: Tokens, Embeddings & Vector Search
    1. 2.1 Token Economics
    2. 2.2 Embeddings — text-embedding-005 vs gemini-embedding-001
    3. 2.3 Firestore Vector Search
    4. 2.4 AlloyDB pgvector & BigQuery
    5. Q&A - 2.1
    6. Q&A - 2.2
    7. Q&A - 2.3
    8. Q&A - 2.4
  3. MODULE 3: Prompt Engineering & Structured Output
    1. 3.1 System Prompts & Generation Config
    2. 3.2 Few-Shot & JSON Schema Enforcement
    3. 3.3 Chain-of-Thought & Model Routing
    4. Q&A - 3.1
    5. Q&A - 3.2
    6. Q&A - 3.3
  4. MODULE 4: Building RAG — Three Architectures
    1. 4.1 Document AI
    2. 4.2 DIY RAG
    3. 4.3 Vertex AI RAG Engine
    4. 4.4 Vertex AI Search + Google Grounding
    5. Q&A - 4.1
    6. Q&A - 4.2
    7. Q&A - 4.3
    8. Q&A - 4.4
  5. MODULE 5: BigQuery ML & SQL-Native AI
    1. 5.1 ML in SQL
    2. 5.2 Time Series & Anomaly Detection
    3. 5.3 LLM in SQL
    4. 5.4 BigQuery → Vertex AI
    5. Q&A - 5.1
    6. Q&A - 5.2
    7. Q&A - 5.3
    8. Q&A - 5.4
  6. MODULE 6: Function Calling & Tool Use
    1. 6.1 Gemini Function Calling
    2. 6.2 Complete Calling Loop
    3. 6.3 Parallel Calls & Built-in Tools
    4. Q&A - 6.1
    5. Q&A - 6.2
    6. Q&A - 6.3
  7. MODULE 7: MCP Servers on Cloud Run
    1. 7.1 Building FastMCP Server
    2. 7.2 Deploy to Cloud Run with IAM
    3. 7.3 Connect Agent to Remote MCP
    4. Q&A - 7.1
    5. Q&A - 7.2
    6. Q&A - 7.3
  8. MODULE 8: AI Agents — ADK, Agent Engine, A2A
    1. 8.1 Root Agent with ADK
    2. 8.2 Multi-Agent Orchestration
    3. 8.3 Agent Engine
    4. 8.4 A2A Protocol
    5. Q&A - 8.1
    6. Q&A - 8.2
    7. Q&A - 8.3
    8. Q&A - 8.4
  9. MODULE 9: Multimodal — Vision, Speech & Media
    1. 9.1 Gemini Multimodal
    2. 9.2 Generative Media
    3. 9.3 Pre-trained APIs
    4. Q&A - 9.1
    5. Q&A - 9.2
    6. Q&A - 9.3
  10. MODULE 10: Fine-Tuning & LLM Optimization
    1. 10.1 SFT with LoRA
    2. 10.2 Context Caching
    3. 10.3 Batch API + Model Routing
    4. 10.4 Vertex AI Evaluation
    5. Q&A - 10.1
    6. Q&A - 10.2
    7. Q&A - 10.3
    8. Q&A - 10.4
  11. MODULE 11: Self-Hosted Models on Cloud Run GPU
    1. 11.1 Deploy Gemma on L4 GPU
    2. 11.2 Custom Inference Endpoint
    3. 11.3 Hybrid Architecture
    4. Q&A - 11.1
    5. Q&A - 11.2
    6. Q&A - 11.3
  12. MODULE 12: Production Deployment & Security
    1. 12.1 Infrastructure & Project Setup
    2. 12.2 RAG API
    3. 12.3 Admin Dashboard
    4. 12.4 Frontend on Cloud Run
    5. Q&A - 12.4
  13. Practice Lab
    1. 1.1 GCP AI Project Setup
    2. 1.2 IAM & Security
    3. 1.3 First Gemini Call
    4. 2.1 Token Economics
    5. 2.2 Embeddings
    6. 2.3 Firestore Vector Search
    7. 2.4 AlloyDB pgvector & BigQuery
    8. 3.1 System Prompts & Generation Config
    9. 3.2 Few-Shot & JSON Schema Enforcement
    10. 3.3 Chain-of-Thought & Model Routing
    11. 4.1 Document AI
    12. 4.2 DIY RAG
    13. 4.3 Vertex AI RAG Engine
    14. 4.4 Vertex AI Search + Google Grounding
    15. 5.1 ML in SQL
    16. 5.2 Time Series & Anomaly Detection
    17. 5.3 LLM in SQL
    18. 5.4 BigQuery → Vertex AI
    19. 6.1 Gemini Function Calling
    20. 6.2 Complete Calling Loop
    21. 6.3 Parallel Calls & Built-in Tools
    22. 7.1 Building FastMCP Server
    23. 7.2 Deploy to Cloud Run with IAM
    24. 7.3 Connect Agent to Remote MCP
    25. 8.1 Root Agent with ADK
    26. 8.2 Multi-Agent Orchestration
    27. 8.3 Agent Engine
    28. 8.4 A2A Protocol
    29. 9.1 Gemini Multimodal
    30. 9.2 Generative Media
    31. 9.3 Pre-trained APIs
    32. 10.1 SFT with LoRA
    33. 10.2 Context Caching
    34. 10.3 Batch API + Model Routing
    35. 10.4 Vertex AI Evaluation
    36. 11.1 Deploy Gemma on L4 GPU
    37. 11.2 Custom Inference Endpoint
    38. 11.3 Hybrid Architecture
    39. 12.1 CI/CD with Cloud Build
    40. 12.4 Frontend on Cloud Run

Questions about this course

How long does it take?

It is self-paced, so it depends on you. There are 121 steps across 12 modules and the practice lab.

Is there a certificate?

No. This course doesn't award a certificate.

What do I need to start?

Module 1 starts by setting up a Google Cloud project for AI, IAM and security, then your first Gemini call. Any usage on your own Google Cloud project is billed by Google, not Netsetos.

bottom of page