I Built a Free AI SEO System That Runs Itself (Exclusive and Free Tutorial)

Back story?
I used to spend hours every week doing the same SEO tasks over and over (not a tech guy). Checking keyword rankings. Running site audits. Pulling data from Search Console. It was boring, time-consuming, and honestly? A waste of my skills.

Then I discovered you could build your own AI assistant to do all of this automatically. No coding required. Completely free (well, mostly), and it actually works.

Here’s how I did it, and how you can build one too.

What This AI Assistant Actually Does

Think of it as having a junior SEO analyst who never sleeps, never complains, and works for free. Here’s what mine can do:

The Main Tasks:

• Finds keywords with search volume and difficulty scores
• Checks where my pages rank in Google
• Pulls traffic data from Search Console
• Submits URLs to Google for indexing
• Runs full website performance audits
• Saves everything to Google Drive automatically

And the best part? I just tell it what I want in plain English. No complicated commands or programming.

The Tools You’ll Need

Here’s everything required to build this. Don’t worry, most of it is free:

Complete Setup (Step by Step Tutorial)

Part 1: Install Google Antigravity (15 minutes)

Step 1: Go to the Google Antigravity website and download the installer for your system (Mac or Windows).

Step 2: Run the installer and follow the prompts. It’s like installing any other app—just click Next a few times.

Step 3: Open Antigravity and sign in with your Google account. Use the same account you use for Gmail and Google Drive.

Step 4: If you have Gemini Pro, connect that account too. If not, the free tier works fine to start.

Part 2: Set Up Google Cloud Console (20 minutes)

This is where you get permission to access Google’s APIs. Don’t worry, it’s free.

Step 1: Go to console.cloud.google.com and sign in.

Step 2: Click “Create Project” at the top. Name it SEO Assistant or whatever you want.

Step 3: Once created, click on Enable APIs and Services.

Step 4: Search for and enable these APIs one by one:

• Google Drive API
• Google Sheets API
• Google Docs API
• Google Search Console API
• Google Indexing API

Step 5: Go to Credentials in the left menu.

Step 6: Click Create Credentials – OAuth client ID

Step 7: Choose Desktop app and give it a name.

Step 8: Download the JSON file. Rename it to credentials.json.

Step 9: Save this file somewhere safe. You’ll need it in the next step.

Part 3: Create Your Project Structure

Step 1: Open Antigravity and click New Project.

Step 2: Create a folder on your computer called SEO Executive.

Step 3: Inside that folder, create a new file called readme.md.

Step 4: Open readme.md in any text editor and paste this:

# SEO Executive AI Assistant

You are an SEO professional assistant. Your job is to:
- Perform keyword research using DataForSEO API
- Check keyword rankings for websites
- Extract data from Google Search Console
- Submit URLs for indexing
- Audit website performance
- Save all results to Google Drive in organized folders

Always ask for clarification if instructions are unclear.
Use batch processing when possible to save time.
Learn from errors and avoid repeating mistakes.

Step 5: Copy your credentials.json file into the SEO Executive folder.

Step 6: In Antigravity, click Load Project and select your SEO Executive folder.

Step 7: Click Initialize Project. The AI will read your readme and set up the folder structure automatically.

Part 4: Set Up DataForSEO

Step 1: Go to dataforseo.com and create a free account.

Step 2: After signing in, go to your Dashboard.

Step 3: Click on API Access or Credentials.

Step 4: Copy your API Login (username) and API Password. Save these somewhere, you’ll need them soon.

Step 5: Note: You get free credits when you sign up. Enough to test everything.

Part 5: Download Location Codes

For location-based keyword research, you need a list of location codes.

Step 1: In Antigravity, tell your AI: Download the DataForSEO location codes CSV file and save it to the project folder.

Step 2: The AI will fetch the file automatically. This contains codes for cities, states, and countries worldwide.

Building Each Workflow (Step-by-Step)

Building the Keyword Research Workflow

Step 1: In Antigravity, type this command to your AI:

Create a keyword research workflow. Use DataForSEO API to get keyword data including search volume, difficulty, and CPC. Save results to Google Sheets and upload to Google Drive.

Step 2: The AI will ask for your DataForSEO credentials. Paste your API Login and Password.

Step 3: Wait 2-5 minutes while it builds the workflow. You’ll see it writing code and testing connections.

Step 4: When it says Workflow complete, test it by saying: Find keywords for ‘best laptops’ in United States.

Step 5: Check your Google Drive. You should see a new spreadsheet with keyword data.

What just happened: The AI created Python scripts, connected to DataForSEO, set up Google Sheets formatting, and tested everything automatically.

Building the Rank Tracking Workflow

Step 1: Tell your AI:

Create a rank tracking workflow. Check keyword rankings for given URLs using DataForSEO. Support multiple URLs and keywords. Save results to Google Sheets.

Step 2: The AI will use your existing DataForSEO credentials.

Step 3: Wait for it to build and test the workflow.

Step 4: Test it: Check where mywebsite.com ranks for best coffee makers’ in New York.

Step 5: Verify the results in the Google Sheet it creates.

Building the Search Console Data Workflow

Step 1: First, make sure your website is verified in Google Search Console.

Step 2: Tell your AI:

Create a workflow to extract performance data from Google Search Console. Get clicks, impressions, CTR, and position for any URL I specify.

Step 3: The first time this runs, a browser window will open asking you to authorize access. Click Allow.

Step 4: Test it: Get Search Console data for mywebsite.com/blog for the last 30 days.

Step 5: The AI will display the data and offer to save it to Drive.

Building the URL Indexing Workflow

This one requires extra setup in Search Console.

Step 1: In Google Cloud Console, create a Service Account:

• Go to IAM & Admin → Service Accounts
• Click Create Service Account
• Name it SEO Indexing Bot
• Copy the service account email (looks like: seo-bot@project.iam.gserviceaccount.com)

Step 2: In Google Search Console:

• Go to Settings → Users and permissions
• Add the service account email as a user with Full” permissions

Step 3: Back in Antigravity, tell your AI:

Create a workflow to submit URLs for indexing using Google Indexing API. Support batch submissions.

Step 4: Test it: Submit mywebsite.com/new-page for indexing.

Step 5: Check Google Search Console to confirm the submission went through.

Building the Website Audit Workflow

Step 1: Tell your AI:

Create a Core Web Vitals audit workflow. Use PageSpeed Insights API and browser automation to analyze pages. Extract performance metrics, identify issues, and provide specific optimization recommendations. Save full reports to Google Docs.

Step 2: Wait while it sets up browser automation and API connections.

Step 3: Test it: Audit mywebsite.com and give me a performance report.

Step 4: The AI will visit the page, run tests, and create a detailed report in Google Docs.

Step 5: Check your Drive for the audit document with scores and recommendations.

Making Your Assistant Faster

After building all workflows, optimize performance:

Step 1: Tell your AI:

Optimize all workflows for speed. Use batch processing for API calls. Store temporary data in variables instead of files. Reduce unnecessary operations.

Step 2: Wait while it reviews and updates each workflow.

Step 3: Test a few commands to see the speed improvement.

Teaching Your Assistant to Learn

Step 1: Tell your AI:

When you encounter errors, analyze them, fix the underlying code, and document the solution so you don't make the same mistake again.

Step 2: From now on, when errors happen, watch the AI troubleshoot and fix itself automatically.

Building Your AI’s Skills (The Coolest Part)

Your AI starts with zero skills. You teach it by describing what you want. Here’s how I added each capability:

Keyword Research

I told my assistant:

Create a workflow that gets keyword data from DataForSEO and saves it to Google Sheets.

It took about 3 minutes, and boom, I had a working keyword research tool. Now when I say Find keywords for ‘cosmetic brand’ it:

• Pulls search volume
• Gets difficulty scores

Here’s what the output looks like:

Rank Tracking

Same process. I said:

Check where my URLs rank for specific keywords.

Now I can give it a list of my pages and keywords, and it checks them all. Results go straight to a Google Sheet.

Example output:

Getting Search Console Data

This one’s super useful. I ask: What’s the traffic for example.com/blog?

The AI pulls clicks, impressions, and CTR from Search Console. No more logging into dashboards manually.

Website Performance Audits

This is probably my favorite. The AI visits my pages, runs performance tests, and gives me a full report with suggestions.

Typical audit results:

The full report goes to Google Docs with screenshots and specific recommendations.

What Makes This Different From Regular SEO Tools

Regular tools: You click buttons, wait for results, export data, organize it yourself.

This AI assistant: You say what you want, it figures out how to do it, and delivers organized results to your Drive.

Plus, it learns. When it hits an error, it fixes its own code and doesn’t make that mistake again. I’ve watched it troubleshoot API issues and optimize its own workflows without me touching anything.

Examples of How I Use It

Monday Morning Routine: Check rankings for all my target keywords and compare to last week. Gets a comparison report in 2 minutes instead of 30.

Client Reporting: Pull Search Console data for client-site.com for the last 30 days. Complete traffic report ready to send.

New Content Launch: Submit these 10 URLs for indexing and check their Core Web Vitals. Everything done while I grab coffee.

Competitor Research: Find keywords that competitor.com ranks for in positions 1-10. Full competitor keyword list in minutes.

The Honest Challenges

It’s not instant: Setting up the first few workflows takes time. Budget 2-5 minutes per new skill.

You need to approve things: When it accesses APIs for the first time, you’ll need to click Allow on permission prompts.

DataForSEO costs money: After free credits run out, you’ll pay for API calls. But it’s cheaper than most SEO tools ($0.20-0.40 per task usually).

Learning curve exists: Not steep, but you’ll spend an afternoon getting everything configured initially.

Making It Faster and Smarter

After I built the basic workflows, I told my assistant:

Make everything faster.

It went through and optimized its own code. Started using batch processing. Reduced unnecessary steps. Now tasks that took 3 minutes run in 30 seconds.

I also said:

Learn from your mistakes and don't repeat them.

It actually does this. Keeps track of errors and adjusts its approach automatically.

What Your AI Assistant Can’t Do (Yet)

Let’s be realistic about limitations:

• Can’t write content for you (that’s still your job)
• Won’t develop your SEO strategy (you need to tell it what to check)
• Doesn’t work with every SEO tool out there (mainly Google and DataForSEO)
• Can’t read your mind (you need to give clear instructions)

But honestly? For repetitive data tasks, it’s a game changer.

Should You Build This?

Build it if you:

• Spend hours on repetitive SEO tasks
• Want to automate reporting
• Manage multiple websites
• Like having organized data in one place
• Are comfortable with a bit of tech setup

Skip it if you:

• Only manage one small site
• Prefer clicking through tools manually
• Don’t want to deal with any API setup
• Need a polished, ready-made solution

Getting Started This Week

Here’s your weekend project plan:

Saturday morning: Set up Antigravity and Google Cloud Console (1 hour)

Saturday afternoon: Build your first workflow, start with Search Console data (1 hour)

Sunday: Add keyword research and rank tracking (2 hours)

Next week: Add more skills as you need them (ongoing)

The Bottom Line

I’m not saying this replaces professional SEO tools entirely. But for the boring, repetitive stuff? It’s perfect.

I’ve saved probably 5-10 hours per week since building this. That’s time I now spend on actual strategy, content planning, and work that requires human creativity.

The best part? Once it’s set up, it just runs. I fire off a quick command in plain English, and my reports are ready before I finish my coffee.

If you’re tired of spending hours on SEO busywork, give this a shot. Yes, there’s setup involved. Yes, you’ll need to fiddle with APIs and credentials. But once it’s running, you’ll wonder how you ever did SEO without it.

Quick Reference: What Each Component Does

Now go build your AI assistant. Your future self (with 10 extra hours per week) will thank you.

FAQs