Step 2: Setting up Google Cloud & Vertex AI
Now that we know our goal (a chatbot powered by LLMs), the first thing we need is an environment to run our AI models. We’ll use Google Cloud Platform (GCP) and its Vertex AI service.
💡 Why Google Cloud & Vertex AI?
- Vertex AI allows us to use Google’s Large Language Models (LLMs) without training them ourselves.
- We can call these models via APIs (similar to ChatGPT) to generate chatbot responses.
- Our local machine (PyCharm/Streamlit app) will send user prompts → Vertex AI → return answers.
- This avoids needing expensive GPUs on our laptop.
✅ What We Need
- A Google Cloud Account (can be a regular Gmail, or a Google Workspace account).
- A Google Cloud Project – this acts like a container for our AI app.
- Enable Vertex AI API inside the project.
- A Service Account + API Key or JSON Credentials to allow our app to talk to Vertex AI securely.
🛠 How to Get What We Need
Step 2.1: Create a Google Cloud Account
- If you already have a Gmail account → you can use it directly.
- If not, create a Gmail account.
- Visit Google Cloud and sign in.
👉 You’ll usually get $300 in free credits when you first sign up.
Step 2.2: Create a New Project
- Go to Google Cloud Console.
- Click the Project Selector at the top → New Project.
- Give it a name like
chatbot-project. - Click Create.
Step 2.3: Enable Vertex AI API
- In the left sidebar, go to APIs & Services → Library.
- Search for
Vertex AI API. - Click Enable.
Step 2.4: Set up Authentication
Our app needs permission to use Vertex AI. For this, we use a Service Account.
- In the Cloud Console, go to IAM & Admin → Service Accounts.
- Click Create Service Account.
- Name it
vertex-chatbot. - Assign the role: Vertex AI User.
- Click Create Key → JSON.
- A
.jsonfile will be downloaded → keep this safe! (we’ll use it in PyCharm).
📂 Save the JSON file in your project folder (e.g.,
~/chatbot/key.json).🚀 What’s Next?
We now have Google Cloud + Vertex AI ready. In the next step, we’ll connect our local Streamlit app to Vertex AI using the API key / JSON credentials and make our first chatbot request.