Integrating N8n with Whisper, ChatGBT and WordPress: A Journey

Apr 23, 2025 | , , , , , , , ,

When I embarked on my journey of getting N8n to work with Whisper, ChatGBT and WordPress, little did I know about the complexity that lay ahead. On the face of it, everything seemed fairly straightforward – after all, there should ideally be a node for it, right? To my surprise, nodes for Whisper and ChatGBT were hard to come by. I’ve heard rumblings about some recent developments in this direction, but they seemingly hadn’t wooed my release channel yet. The fast-paced evolution of technology throws us such curveballs, doesn’t it?

My aim was to explore the exciting domain of agentic AI. The idea was not just to have the AI supplying code samples or engaging in a casual back-and-forth. I wanted it to actually ‘do’ something for me. And I must say, the mission was a success. The AI took commands about my interests and needs, and transformed them into a blog post in language that felt natural and less AI-ish.

However, this wasn’t without its fair share of hurdles. N8n’s UI came with some stringent limitations. It didn’t give me options to send the binary data of the recording along with the parameters for the Whisper model. On top of this, navigation was a major pain since the UI was virtually inaccessible to screen-readers — a frustratingly counterintuitive experience for the most part. With essential functions buried behind unlabeled buttons, it was like finding a needle in a haystack.

To counter these obstacles, I did some sleuthing into the JSON for the node I was writing. Once figured out, this could be directly uploaded to the workflow for testing. Again, testing was not a cake-walk given that the test button wasn’t always an available choice. Sometimes, I had to run the entire workflow to test it, which as you can imagine, was quite taxing.

Besides, the error-check button wasn’t exposed via ARIA. By the way, ARIA, or Accessible Rich Internet Applications, is a standard that aids screen-readers in understanding HTML objects that lack text alternatives. So there you have it, another bump in the road.

Despite these setbacks, I persisted and was finally able to demonstrate the sheer potential of N8n, proving it as a truly powerful tool. Mind you, someone with a lesser technical prowess might find this a tough nut to crack. For instance, to properly encode the body parameter for Whisper, I had to write my own server. Additionally, I had to write JavaScript to marshal the JSON sent from ChatGPT, because everything came through packed in the content field. To allow mapping of attributes to the inbound parameters of the next node, creating a new JSON object that explicitly provided top-level JSON objects was necessary.

In a nutshell, while this journey was fraught with challenges, it was rewarding in that it made me realize the immense potential of agentic AI in action. Do bear in mind, however, that navigating through this terrain requires a solid skillset, along with heaps of patience, resilience, and problem-solving ability.

Harnessing the Power of N8n and ChatGPT for Enhanced Content Automation

Apr 23, 2025 | , , , , ,

Today has been filled with riveting experiences of content automation, and I would be delighted to walk you through the highs of my day. I had the opportunity to experiment with N8n, an incredible tool, to integrate Faradite on my iPhone – which was an intriguing endeavour indeed.

Now if you’re wondering what all these terms mean, allow me to elaborate. N8n is a free and open workflow automation tool. It allows for the intercommunication and integration of various applications through a concept known as webhooks: HTTP push API from apps. With this capacity, I used HTTP webhook to post the audio.

Faradite is an application on my iPhone that I utilised in this automation endeavour. Now, from Faradite, the integrated data in the form of audio input went straight to OpenAI or ChatGPT Whisper.

ChatGPT, another AI-powered software tool, processed this data and accomplished a fascinating feat—it transcribed the audio into textual content—turning spoken words into written ones. What an exceptional way to create content, isn’t it?

This tactic of integrating modern tech tools to create more sophisticated and seamless workflows promises not just convenience, but high-quality results as well. I eagerly look forward to seeing the transcribed outcome of this first post, given that it’s quite late, and this could possibly be classified as a late-night ideation ramble.

ChatGPT now has the baton in its court, to transform this audio recording into an engaging blog post. I can’t express enough how remarkable this entire process is, opening doors to more innovative ways of content creation in future.

Using a Mac as my daily driver.

Oct 16, 2024 | , ,

Well, this could be interesting. 

About five days ago, I purchased a Mac Book Pro M3 Max.  Shed lodes of RAM, CPU cores and GPU cores as well as a 1TB disk. 

I’ve had a few Mac machines over the years but they have always been secondary machines. 

About nine months ago I began using a Mac more and more. Primarily for sound editing.  

So decided to take the plunge. 

But I can’t really justify paying all this money for a machine that has a single purpose. 

So I’m making myself try to use this as my main machine. 

Problem is, I’m working almost exclusively within the Microsoft ecosystem. So this isn’t without its problems. 

 

I’ll let you know how this goes. 

VSCode indentation navigation made simple.

Dec 20, 2023 | , , , , ,

It could be PowerShell, Python or Yaml. It doesn’t really matter too much. Some languages obviously require indents for code structures where others require brackets but coding in any kind of collaborative environment requires that your code is readable therefore indents are required regardless. So navigating through code based on indentation level is particularly important and useful.

I was trying to find a way of making my prefered screen reader notify me when indentation levels changed when I stumbled on a handy extension for VSCode called “Indent Nav“.

I’ll explain why I love it with a very relatable example:

I’ve written some code but I’ve decided eventually that this code should be contained within an if block or I might even want to move it into it’s own function.

I move to the first line of the block of code. As it happens, I’m checking for a condition so it’s an if block. I press F9 to start selection. Then I hit alt end to move to the end of that block and I press alt f10 to mark the end of the selection. Job done! I hit control C, type my new if block or create the function and I paste. This is a simple example but it saves vital time.

Or the other example is I’m using splatting. The splat is long. I could arrow through every line while reading the code but I don’t need to. I press alt and down to move to the next line at the same indent level and I skip it all.

Yes, I could use code folding to skip code blocks that I don’t need to read but sometimes out of sight out of mind is not a good thing. I like to have the code there on screen without using folding unless it’s at the function level as it’s very easy to miss vital clues when debugging if your cursor automatically skips over blocks of code. That’s probably jsust me though.

Chat GPT Saved Me Time and Effort in Building a Dynamic Country Dropdown in Blazor

May 15, 2023 | , ,

Introduction:

Have you ever found yourself drowning in a sea of code, desperately trying to build a seemingly simple feature for your application? That’s exactly where I was, until I discovered a superhero in the form of Chat GPT. In this post, I’ll regale you with the thrilling tale of how Chat GPT swept in, saved the day, and helped me build a dynamic country dropdown in Blazor—all while keeping my sanity intact. So, hold onto your coding capes and get ready for a wild ride!

The Dilemma:

It all started when I needed to implement a dynamic country dropdown in my Blazor application. I went looking for a CSV file containing a comprehensive list of countries and their corresponding codes. The task seemed simple at first glance, but as I delved deeper, the enormity of the challenge became apparent. Manually processing the CSV file, creating a C# class, and integrating it with a Blazor form component seemed like a daunting and time-consuming task. Who wants to do all that messing around for hundreds of countries?

Enter Chat GPT, My AI Sidekick

That’s when Chat GPT came to my rescue like a coding wizard in a digital cape. With its natural language understanding and coding capabilities, I quickly realized that it could handle the entire process for me. All I had to do was provide the CSV file and outline the requirements, and Chat GPT would take care of the rest. It was like having a knowledgeable coding buddy by my side.

Processing the Country List:

I started by feeding the CSV file to Chat GPT. It quickly parsed the data and generated a C# class with the necessary fields for country names and codes. This automated process saved me hours of manual data entry and ensured accuracy.

Crafting the Blazor Form Component:

Next up was the Blazor form component. Armed with the C# class generated by Chat GPT, I was able to seamlessly integrate it into my application. The country dropdown was dynamically populated with the countries and codes, all thanks to the magic of Chat GPT. It even provided me with sample code snippets to handle the selection and retrieval of the chosen country.

JavaScript Enchantment:

To ensure the accessibility of the dropdown for screen reader users, I needed a touch of JavaScript wizardry. And who would have thunk it? Chat GPT had just the spell for it. It offered me JavaScript interop methods to retrieve elements by their enchanting IDs and set focus with a wave of its metaphorical wand. With a sprinkle of JavaScript magic, I transformed the dropdown into a user-friendly and inclusive masterpiece. well, maybe not. That’s a bit over board, but damn, I was happy with it!

The Grand Finale: Empowered by AI:

What could have been an arduous task turned into a joyous coding adventure. By harnessing the power of AI, I unlocked new levels of productivity and creativity. The ability to delegate repetitive and time-consuming tasks to AI allowed me to focus on higher-level thinking and problem-solving.

Closing Thoughts: The Unseen Power of AI:

This thrilling coding adventure with Chat GPT taught me a valuable lesson: AI can be a powerful ally in our coding journey. By automating mundane tasks, it frees us to explore new possibilities, think outside the box, and create more efficiently. With the right tools and proper utilization, AI can truly empower us to achieve more, better.

Note: While Chat GPT played a significant role in streamlining my development process, it’s important to mention that human understanding, creativity, and intuition are irreplaceable in software development. AI should be seen as a supportive tool that enhances our abilities, rather than a replacement.

Changing the domain everywhere on a WordPress website. Especially useful for Divi.

Jan 23, 2023 | , ,

Problem

You have a WordPress website but the domain eventually changes for some reason.

You change the domain name in the general settings page but for some reason, you find tha tnot all your iamges are appearing.

You look in the source and easily notice that there are a lot of specific links to your old domain name. Primarily from Dibi.

This isn’t all that easy to fix. For some reason, Divi installs files with the FQDN of your website.

Also, you may have a lot of posts with the old domain name.

So here’s how you can fix all that:

Solution

You will need file system level access and database acccess on the command line to achieve this in the most efficient way. If you don’t have this, perhaps ask your hosting company. I’m sure they would love the extra work…

Alright. I assume you are logged into the MySQL server using the MySQL command line. I’m also assuing you have moved into the correct database using the “use” command.

Must I say it? Back everything up.

Okay. use these commands to replace the domain name in the options, posts and postmeta tables:

update wp_options set option_value = replace(option_value, ‘oldDomain.com’, ‘newDomain.com’);

update wp_posts set post_content = replace(post_content, ‘oldDomain.com’, ‘newDomain.com’);

update wp_postmeta set meta_value = replace(meta_value, ‘oldDomain.com’, ‘newDomain.com’);

Okay. You’re done with the database. Exit out of that back to the command line.

Now update hte files in your website directory as follows: Again, you will need to update the path to your website here.

find /var/www/oldDomain.com/web ( -type d -name .git -prune ) -o -type f -print0 | xargs -0 sed -i ‘s/www.oldDomain.com/www.newDomain.com/g’

Escaping the cloud part 1 – Music

Jan 9, 2023 | , , ,

The cost of living is increasing. So I’m tightening my belt. Are you? Let me know what changes you are making.

One of the changes I have made lately is to remove my dependence on cloud services such as Spotify, iCloud, Google drive etc. There’s a handy feature in my online banking interface that highlights discressionary spending habits. It pointed out about six months ago that I was spending quite a bit for the family on cloud services. So I decided right then and there to fix this madness.

So today is about music. Why am I paying spotify every month? Why am I paying for Apple iCloud and music when there’s such major overlap here? It was just an organic waste that I let sit for far too long. So. What do I really need from a music service?

  • Always accessible
  • Must work with native IOS apps
  • Must work with Sonos
  • Must have a great search function
  • Must let me listen to all tracks at random
  • Must have a nice UI that my wife and children can use
  • Nice to have is multi user access and listening history
  • It also needs to have the usual thing of cover art, playlists, staring / ranking and all the other things you expect from a music service.

Down side: There were some artists that I listened to on Spotify and / or Apple music and I hadn’t yet bought their albums. I’m rather ashamed of this actually. I always make a point of purchasing music from artists that I enjoy. But I had gotten out of the habit lately. so I spent about two hundred Euro on music when I was making this change. I know what you’re going to say. This went completely against my aim to save money. But I promise you dear reader, I would have purchased that music anyway. At some point.

I tried a few services. Navidrome probably came the closest to meeting all my needs but it had one huge flaw. It didn’t support randomly shuffling all my tracks. I might be a bit strange, but I like to just sit down at my desk in the morning and let the music go where ever it wants. One moment it’s playing a slow air. Next it’s playing a mad fast tune with lots of energy. I enjoy it this way.

I settled on Airsonic advanced. It supports Sonos with some messing about and it fits the bill in every other way. It’s not often updated which will probably become a problem but there are a few other projects in the works out there that will hopefully meet my requirements in a year or two.

I had a few challenges:

  • I want my music to be stored on the NAS in my home. Not on the server.
  • My music has a lot of accented characters. I had major problems with mounting the Cifs file system and having those accented characters displayed correctly. I ended up writing a little bit of PowerShell that I ran from my main workstation to just clean up the accented characters. I will paste that below for anyone who wants it.
  • When I add new music, I need to run that script again. I also need to restart the docker container. Just because of the way I configured the mount. It’s probably not the most optomal solution. But it works and it’s really fast.
  • The Sonos configuration is a bit fiddly. But it’s not imporssible. Just be really careful when setting it up.

Other than that, it was all easy.

In the office. Getting less done.

Oct 18, 2022 |

This is absolutely nuts. I’m here writing this blog post a little after 11:30am because it’s something I can do without thinking too much. I’m in the middle of two very complex bits of work that require my full focus and consintration. But I can’t do what I need to because of the noise outside my window and inside the office. Currently, there are two people standing having a chat outside and there’s a very noisy saw running for the past ten minutes. Just before the saw there was someone talking very loudly in the office and there were two other coleagues trying to investigate a 2FA issue. These are all tasks that these people need to cary out and I absolutely don’t have any problem with what all these people are doing. It’s just that I simply can’t focus properly on these tasks. Just when I think the noise is going to stop, it comes back again. Like a knife through my ears.

I read a computer interface using a screen reader that provides feedback using synthesized speech through head phones that I wear constantly. So imagine this from my perspective. It would be like if you were trying to read complicated text off a page but I kept moving the page constantly and continually strobed the lights in the room. It would be very distracting and stressfull. Now imagine I stop for less than a minute and without warning began doing the same thing again. You will be on edge just waiting for the next distraction.

But this is only one part of the reason that this return to the office has me driven round the bend. This morning I left the house at just before 7:50am. The bus didn’t actually arrive until 8:20am and then we got stuck on the M1, the main road between Drogheda and Dublin for ages as a result of a crash. So I didn’t get to the office until 9:40am. If I was working from home, I would start at around 9am but I would have had a good long walk with the dog so my head will be in the right space to have a productive day. I get straight in and with my custom built office environment at home, I can be assured of absolute solitude and quiet.

It’s not all about work as well. At about 3:30pm, I’ll go in to the house and for about 10 minutes, I’ll catch up with my two children to ask them how school was. But I go back to my office again and I’ll easily work until 6pm without any problems. Because I know that when I finish, I have a 1 minute walk back into the house. Where as in Dublin, I’ll finish at 5pm on the dot and I know I have just over an hour before I get home. So there’s no insentive for me to continue working to get the job done.

It just seems sensless to me. My job is not about constantly engaging with people. I value my coleagues and I enjoy collaborating with them. But I find that I can do this equally well remotely. I can work more effectively when I’m not not in the office.

Rant over.

That damn saw is still running. I don’t know how I’m going to get anything done today.

Updating certificates on RDS. Remote Desktop Services

Jul 11, 2022 | , ,

Every year now, I need to update the certificates on my Microsoft Remote desktop services servers.

This involves:

  • IIS front end
  • RDWeb Web client
  • components of RDS through server manager. Connection broker, gateway and web.
  • RDS gateway.

Rough instructions:

Install the certificate

  1. Open the MMC
  2. Click add snap in
  3. Choose Certificates
  4. Choose “Local computer”
  5. Choose this computer
  6. Expand Personal\Certificates
  7. Right click certificates and under all tasks, choose import certificate.
  8. Now import your new PFX file
  9. I recommend giving it a friendly name.
  10. Now right click this certificate and click more tasks.
  11. Click Export
  12. Follow the wizzard. Don’t export the private key.
  13. Save it to somewhere that will be easy to find shortly.

Update IIS

  1. Open IIS Internet Information Services
  2. Expand your server then expand sites
  3. Right click on Default website.
  4. Click Edit bindings
  5. Click on the HTTPS port 443 binding.
  6. Click Edit
  7. Choose your certificate using the friendly name that you configured earlier.
  8. Click Ok then Close.
  9. You can now close the IIS administration interface.

Update the RDWeb web client

You do this by unpublishing the client, importing the certificate then re-publishing the client again.

  1. Open PowerShell as administrator
  2. Type the following command to unpublish the web client.
    Publish-RDWebClientPackage -Type Production
  3. Now import the certificate using the following command. Replace everything between the <> with the path to the cer file you exported earlier.
    Import-RDWebClientBrokerCert
    Publish-RDWebClientPackage -Type Production -latest

Update the RDS service using server manager

  1. Open server management
  2. On the right, open RDS Remote Desktop services
  3. In your main deployment window, click the deployment tasks button.
  4. Click Deployment properties
  5. Highlight the certificates option on the left
  6. For each certificate, do the following:
  7. Click the certificate
  8. Click change
  9. Choose the second option. Ad dfrom file
  10. Browse to the PFX file
  11. Type the password
  12. Click OK
  13. Click Apply
  14. You will need to do this at least four times.

Update the certificate on the gateway

  1. Open the gateway manager
  2. Right click the gateway on the left
  3. Click Properties
  4. Move to the Certificates tab
  5. Choose the third button down to import a new certificate.
  6. Browse to the PFX
  7. Type the password when prompted

At this point, you will probably need to reboot the connection broker and front end servers.

Create a fresh shuffled playlist on your NAS every day.

Jul 8, 2022 | , , ,

Here’s what I needed:

I have a huge amount of music and for some reason, I just like it to play shuffled. When I hit play on the Sonos, I just want it to randemly go through and play the music on the NAS. This actually works beautifully on the Sonos. It doesn’t really need this. But I@m working on something interesting using HomeAssistant at the moment and it doesn’t support the simple requirement of playing all music in a folder. IT requires a playlist.

I already have a Linux server so each day at 3am, when everyone is in bed, a script will run to recreate and re-shuffle the playlist. So when new music is added, it will get added to the playlist and the order is always different.

first of all, put this in /etc/cron.daily but obviously, replace the parts in [] brackets to match your own values.

Okay. you have that file created?

Now set permissions on it using chmod 755.

Finally, add something to your crontab for each day at 3am. First, open crontab by typing crontab -e then press enter.

That’s all there is too it.