Why I Built My Own AI SEO Analysis Tool for Mac
Search engine optimisation has changed considerably over the years. While the fundamentals of good SEO remain important, the way search engines interpret content and the growing influence of artificial intelligence have introduced new challenges for website owners and SEO professionals.
I've worked with websites and search engine optimisation for many years, using a variety of tools to analyse website performance, identify technical problems and improve search visibility.
Recently, I decided to take a slightly different approach: build my own AI-powered SEO analysis application for macOS.
Not because existing SEO software is inadequate, but because I wanted something that approached website analysis from a different perspective.
DOWNLOAD FOR FREE
SEO Companion is available to download and use free of charge. By downloading, installing or using the software, you acknowledge that you have read and agree to the Terms and Conditions.
The software is provided “as is”, without warranties of any kind. South Devon Web Design accepts no responsibility or liability for any loss, damage, data loss or other issues arising from its installation or use, to the fullest extent permitted by applicable law. Technical support and future updates are not guaranteed.

Why Build Another SEO Tool?
There are already some excellent SEO applications available. I regularly use established software to examine website structure, identify technical issues, monitor rankings and investigate potential improvements.
However, traditional SEO tools tend to concentrate on measurable technical factors, such as missing page titles, broken links, heading structures and other established optimisation signals.
These checks are valuable, but they don't necessarily tell the whole story.
A page might pass numerous technical SEO checks while still failing to explain its subject clearly, answer important questions or provide the information visitors are actually looking for.
I wanted to explore whether locally running artificial intelligence could help bridge that gap.
The idea was to develop an application that could examine individual web pages, assess their content and structure, and produce useful recommendations that went beyond a conventional technical checklist.
What Can SEO Companion Actually Do?
SEO Companion is a native macOS application I developed to bring together traditional SEO analysis and locally running artificial intelligence. Rather than replacing established SEO tools, it is designed to complement them, providing a practical way to examine individual web pages, identify potential problems and generate useful recommendations.
The application uses Ollama with the Qwen3:8b AI model, allowing AI analysis to run locally on your Mac without relying on a cloud-based AI subscription.
AI-Powered Content Analysis
SEO Companion examines website content and uses artificial intelligence to identify opportunities for improvement. This includes assessing how clearly a page communicates its subject, whether its content is relevant to its intended search terms and where additional information might improve its usefulness.
The aim is to provide meaningful recommendations rather than simply assigning an arbitrary SEO score.
Traditional SEO Checks
Alongside AI analysis, SEO Companion performs conventional on-page SEO checks, including:
Page titles and meta descriptions.
Heading structure and content organisation.
On-page content and potential optimisation opportunities.
Structured data implementation, including JSON-LD, Microdata and RDFa.
Combining these checks with AI analysis helps identify both technical issues and opportunities to improve the quality and relevance of the content.
Structured Data and Schema Analysis
Structured data is increasingly important for helping search engines understand the information presented on a website.
SEO Companion examines existing schema markup, including JSON-LD, Microdata and RDFa, to help identify potential issues and opportunities for improvement.
This is particularly useful when reviewing pages that describe businesses, services, products or articles, where correctly implemented structured data can provide additional context to search engines.
Website Projects and Multiple Page Analysis
Rather than treating every page as an isolated exercise, SEO Companion allows you to create and save individual website projects.
Pages can be manually added to a project and analysed individually, with the option to review findings across multiple selected pages.
This makes it possible to build a broader picture of a website's SEO performance while keeping each client's information organised.
SEO Recommendations and Progress Tracking
The application generates recommendations based on its analysis, helping identify areas that may require attention.
Recommendations can be managed using To Do and Completed statuses, making it easier to keep track of improvements as work progresses.
These project management features are intended for internal use, keeping the analysis and subsequent optimisation work together in one application.
Exportable SEO Reports
SEO Companion can export its analysis into reports suitable for reviewing or sharing with clients.
Importantly, internal project management information, such as task statuses, completion dates and personal To Do notes, is excluded from client-facing reports by default.
This allows the reports to concentrate on SEO findings and recommendations without including the working notes used to manage the project.
Local AI Processing and Privacy
One of the main reasons I developed SEO Companion was to experiment with running AI locally rather than depending entirely on external services.
By using Ollama and the Qwen3:8b model, the AI analysis runs directly on compatible Apple Silicon Macs.
This removes the need to send the content being analysed to a cloud AI service for processing and avoids ongoing AI API subscription costs.
It also gives me greater control over how the analysis is performed, although the quality of AI recommendations still depends on the model and the information supplied.
A Practical Addition to My SEO Workflow
SEO Companion isn't intended to replace comprehensive SEO platforms, keyword research tools or technical website crawlers.
Instead, it provides another perspective on website content and optimisation, combining conventional SEO checks with AI-generated observations.
It remains an evolving project, developed primarily to support my own website design and SEO work. As I continue using it on real websites, I expect to refine its analysis, reporting and functionality.
For now, it provides a useful demonstration of how locally running AI can be incorporated into a practical SEO workflow without depending on expensive cloud-based services.
Combining Traditional SEO with Artificial Intelligence
My application brings together several different aspects of website analysis.
It examines page content, headings, metadata and structured data, including JSON-LD, Microdata and RDFa.
It also uses AI to help evaluate the information presented on a page and identify potential areas for improvement.
The intention isn't to generate an arbitrary SEO score or suggest changes simply for the sake of making changes.
Instead, I want the application to provide practical observations that I can evaluate using my own SEO experience.
This is an important distinction. AI should assist professional judgement, not replace it.
An AI recommendation isn't automatically correct, and not every suggested change will improve a website's search performance.
Why Run AI Locally on a Mac?
One of the more interesting aspects of this project has been experimenting with locally hosted AI.
Rather than depending entirely on external cloud-based services, the application can use AI models running on the Mac itself.
This approach has several potential advantages, including greater control over the analysis environment, reduced reliance on external services and the ability to experiment with different models.
It also presents its own challenges. Local AI performance depends on the hardware available, the model being used and the complexity of the analysis.
For me, this has been as much a software development and learning exercise as an SEO project.
Developing the Application
What started as a relatively simple experiment has gradually evolved into a more complete macOS application.
I've introduced the ability to create and save individual website projects, add pages for analysis, review recommendations and export reports.
I've also been developing the reporting system to make the results useful in a real-world SEO workflow.
One particularly important consideration has been separating internal working notes from information intended for clients.
An SEO report should explain the issues identified and the improvements being recommended without overwhelming the reader with unnecessary technical information.
There have been plenty of revisions, unexpected problems and changes of direction along the way, which is very much part of developing software.
Will AI Replace Traditional SEO Tools?
I don't believe so.
Website crawlers, ranking software, performance testing and search analytics all provide information that an AI content analysis tool cannot simply replace.
Google Search Console, for example, provides actual search performance data. AI analysis does not.
My application is designed to complement these tools, offering another perspective on website content and optimisation opportunities.
And while AI is becoming increasingly prominent in search, I still believe that clear, useful, well-structured content remains fundamental to building a successful website.
What Happens Next?
The application is still being developed, and I'm continuing to refine the interface, analysis process and reporting features.
I'm also considering making a version available as a free download so that other website owners and SEO professionals can experiment with it and provide feedback.
For now, it's an interesting example of how traditional SEO experience and newer AI technologies can work together.
Perhaps more importantly, building the application has encouraged me to question some of the assumptions behind conventional SEO analysis.
And that, in itself, has made the project worthwhile.
Download for Free
The AI SEO Analysis Tool is provided free of charge by South Devon Web Design on an “as is” and “as available” basis, without warranties or guarantees of any kind, to the fullest extent permitted by applicable law.
While reasonable care has been taken during development, no guarantee is made regarding the application's accuracy, reliability, compatibility, performance or suitability for any particular purpose.
South Devon Web Design accepts no obligation to provide technical support, maintenance, updates or bug fixes.
Users install and operate the software at their own risk and are responsible for reviewing any SEO recommendations before implementing changes to their websites.
To the fullest extent permitted by law, South Devon Web Design accepts no liability for any loss, damage, data loss, business interruption or other issues arising from the installation or use of the application.
Nothing in this disclaimer excludes or limits liability that cannot lawfully be excluded under UK law.
Frequently Asked Questions About SEO Companion
Is SEO Companion completely free?
Yes. SEO Companion is available as a free download, with no subscription fees or ongoing licence costs. The application uses locally running AI, so there are no additional AI API charges.
The software is provided free of charge and subject to the published Terms and Conditions.
Does SEO Companion require an internet connection?
SEO Companion uses Ollama to run its AI model locally on your Mac. Once Ollama and the required model have been installed, the AI processing itself does not require an internet connection.
However, an internet connection is required to download the software, install the AI model and retrieve live website pages for analysis.
What are the system requirements?
SEO Companion has been developed and tested on Apple Silicon Macs running macOS.
You will also need Ollama installed, together with the Qwen3:8b language model. The model requires several gigabytes of storage and sufficient available memory to operate effectively.
Performance will depend on your Mac's specifications.
How do I download and install SEO Companion?
SEO Companion is distributed as a ZIP file, available through the download link on this page.
Download the ZIP file, extract its contents and move the application into your Mac's Applications folder.
You will also need to install Ollama and download the Qwen3:8b model before using the AI analysis functionality.
Because the application is independently developed, macOS may display a security warning when you first attempt to open it. Follow the installation instructions on this page to allow the application to run.
Does SEO Companion send my website content to cloud-based AI services?
No. SEO Companion uses Ollama to process AI analysis locally on your Mac rather than sending website content to an external AI provider.
This provides greater control over the analysis process and avoids cloud AI processing charges.
The application still needs to retrieve publicly accessible website content when performing its analysis.
What SEO checks does the application perform?
SEO Companion combines conventional on-page SEO checks with AI-assisted content analysis.
It can examine page titles, meta descriptions, heading structures, website content and structured data implementation.
The AI analysis also identifies potential content gaps, evaluates how clearly a page communicates its subject and suggests opportunities for improvement.
Can SEO Companion analyse structured data and schema markup?
Yes. The application examines three common structured data formats: JSON-LD, Microdata and RDFa.
It can identify existing structured data and highlight potential inconsistencies between schema descriptions and the visible page content.
However, it is not a replacement for dedicated schema validation tools or Google's Rich Results Test.
Can I analyse an entire website?
SEO Companion allows you to create website projects and manually add individual pages for analysis.
You can examine pages separately or review findings across multiple selected pages to identify broader content and optimisation opportunities.
It is not intended to replace a full website crawler.
Can I export SEO reports for clients?
Yes. SEO Companion can generate exportable SEO reports containing analysis findings and recommendations.
Internal project management information, including To Do statuses, completion dates and personal working notes, is excluded from client-facing reports by default.
This helps keep exported reports focused on the information relevant to the website owner.
How does SEO Companion compare with commercial SEO tools?
SEO Companion is designed to complement established SEO applications rather than replace them.
Commercial platforms often provide extensive functionality such as keyword tracking, backlink analysis, competitor research and automated website crawling.
SEO Companion concentrates on individual page analysis, structured data, AI-assisted content recommendations and project-based reporting.
Its main advantage is the ability to combine conventional SEO checks with locally running artificial intelligence without ongoing AI subscription costs.
Is technical support available?
SEO Companion is an independently developed application provided free of charge.
It is supplied on an "as is" basis, without guaranteed technical support, maintenance or future updates.
Although I may continue improving the application, I cannot guarantee compatibility with every Mac configuration or provide individual troubleshooting assistance.
Please read the Terms and Conditions before downloading or installing the software.
Will using SEO Companion improve my Google rankings?
SEO Companion can help identify technical issues, content gaps and opportunities to improve a website's search engine optimisation.
However, no SEO application can guarantee higher rankings, increased traffic or inclusion in AI-generated search results.
The recommendations should be reviewed and applied using appropriate SEO knowledge and judgement. The AI analysis is intended to assist decision-making, not replace it.