Tool Setup Guide

Google AI Studio + FlowiseAI

Goal: This page helps you get ready to build agentic AI workflows using Google AI Studio and FlowiseAI. Setup is separated from the home page so the home page can focus on topics only.
Quick Path (Most Students)
  1. Google AI Studio: open it in your browser and confirm you can run a sample prompt.
  2. Docker Desktop: install and confirm it is running.
  3. FlowiseAI: run the Docker image flowiseai/flowise, map port 3000, and open http://localhost:3000.

Expected time: ~15–25 minutes (including Docker install).

Page Navigation

A) Official Resources (Links)

Back to top

1) Google AI Studio (Browser Tool)

What you do in this course
  • Experiment with prompts and agent behavior design
  • Test tool-use ideas and multi-step reasoning
  • Document prompts/results for reproducibility
Setup Steps
  1. Open: Google AI Studio
  2. Sign in with your Google account.
  3. Run a simple prompt (example): 
    Summarize what makes a prompt different from an agent in 3 bullets.
  4. If you get a response, you are done.
Important: Do not paste private, sensitive, or confidential information into AI tools.

Back to top

2) FlowiseAI (Docker Desktop Method)

Course rule: Do not install FlowiseAI via NodeJS/npm for this course. We use Docker to avoid environment and dependency issues.
Why Docker?
  • No NodeJS/npm installation required
  • Same environment for everyone (students + TAs)
  • Fewer setup failures and faster support
Step A — Install Docker Desktop
  1. Install Docker Desktop: Docker Desktop docs
  2. Launch Docker Desktop and confirm it is running.
  3. Restart your computer if prompted.
Step B — Pull and Run FlowiseAI (Docker Desktop UI)
  1. Open Docker Desktop
  2. Go to Images
  3. Search: 
    flowiseai/flowise
  4. Click Pull
  5. Click Run

Required Run Settings

  • Ports
    • Host port: 3000
    • Container port: 3000
  • Volume (Strongly Recommended – keeps your flows saved)
    • Container path: 
      /root/.flowise
    • Volume name: 
      flowise_data
Step C — Verify FlowiseAI Works

Open your browser and go to:

http://localhost:3000

If the Flowise UI loads, your setup is complete.

Stop/Restart

  • Docker Desktop → Containers → Stop / Start

Back to top

3) YouTube Walkthrough (Recommended Video)

Recommended (Docker-based) installation video

Watch this video for a visual walkthrough of the Docker-based setup:

If the video differs slightly from your Docker Desktop UI, follow the written instructions on this page.
Not supported for this course (npm/NodeJS)

This video installs Flowise via NodeJS/npm. It is not supported in this course due to frequent setup failures:

Back to top

4) Completion Checklist (Before Labs)

You are ready when:
  • Google AI Studio opens and you can run a test prompt
  • Docker Desktop is installed and running
  • Flowise container is running
  • http://localhost:3000 loads the Flowise UI

Back to top

5) Common Issues (Student-Friendly)

Flowise does not load
  • Confirm Docker Desktop is running
  • Confirm the Flowise container is running
  • Confirm port mapping is 3000 → 3000
Port 3000 is already in use
  • Stop other apps using port 3000 (local servers, previous containers)
  • Restart Docker Desktop
  • Try running Flowise again
My flows disappeared
  • You likely did not set the volume mapping.
  • Re-run the container with volume:
Container path: /root/.flowise
Volume name: flowise_data

Back to top

TA / Instructor Notes

Support boundaries
  • If a student installed Flowise via npm/NodeJS, do not debug it—redirect to Docker method.
  • Fast reset: stop and delete the Flowise container, then re-run with correct ports/volume.
  • If Docker cannot be installed (managed laptop), assign a hosted/shared instance if available.

Back to Home



© 2026 by Dr. Chih Hao Ku at chih-hao.ku@unt.edu