5 practical tools I built with AI.

AI changed the distance between “I wish there was a tool for this” and “I can build this myself.”

As a web designer, I regularly run into small repetitive tasks: cropping transparent images, changing opacity across many files, creating video thumbnails, preparing images for SEO or removing a background.

None of these problems are especially exciting. But they cost time.

So instead of adding another SaaS subscription to my workflow, I started building focused browser tools with the help of AI.

The result is a growing collection at lnny.de/tools.

One problem.
One tool.
As little friction as possible.

Five small utilities I built around the kind of repetitive work that appears constantly in design, development and SEO.

01

Image Processing

Bulk crop transparent PNGs and normalize them onto a consistent canvas.

Transparent padding sounds trivial until you have thirty product images, icons or logos that all sit differently inside their files.

PNG Sculptor removes transparent borders and lets me place the results onto a consistent output size with optional padding. The real benefit is not cropping — it is visual consistency.

02

Image Processing

Apply one opacity value to many images without opening a full design application.

Sometimes the task really is just: make all of these images 40% transparent.

Opacity Studio lets me upload several images, apply a global opacity value and export the processed files. Everything happens directly in the browser.

03

AI Image Tool

Remove an image background and export a transparent PNG — with a deliberately simple workflow.

Upload an image, remove the background, optionally crop the result and download it.

I like this kind of AI integration because the AI itself is almost invisible. There is no chatbot and no complicated prompt layer. It simply solves one specific problem.

04

SEO

Turn inconsistent image filenames into a structured, SEO-friendly naming workflow.

A filename like IMG_5829.jpg says almost nothing. A structured filename can include useful context around the subject, category, brand or dimensions.

For ten images, manual naming is fine. For hundreds of assets, consistency matters — and repetitive work is exactly where automation becomes valuable.

05

Video

Pick a frame from a video and turn it into a polished poster image for the web.

Creating a poster image manually usually means opening a video, finding the right frame, exporting it, resizing it and potentially adding a play button.

Video Poster Pro compresses that into one browser workflow. Choose the frame, choose the play-button style and export the result.

Why build them?

There are already thousands of online tools. The problem is that a tiny task can easily become a whole product experience: create an account, confirm an email, start a trial, navigate a dashboard and discover that the useful export option sits behind a subscription.

A task that should take 20 seconds should feel like a 20-second task.

That is the reason I like focused tools. They do not need to become startups. They do not need accounts, dashboards or onboarding.

They simply need to solve one problem well.

AI + Development

All of these tools were created with the help of AI. But built with AI does not automatically mean AI-powered.

Some tools may use machine-learning technology as part of their actual functionality. Others are traditional browser applications whose development was accelerated by AI.

For me, AI behaves more like a technical collaborator. I can start with a problem, describe the workflow, build an initial version, test it, discover edge cases and refine it quickly.

That changes what is worth building. Small utilities that would previously have required too much setup can now become useful working products surprisingly fast.

The important part

AI can generate code quickly. That does not mean every generated solution is good.

I still have to make decisions about user experience, interface design, performance, browser behavior, responsive layouts, SEO, accessibility, privacy and error handling.

AI can accelerate implementation. It cannot decide whether the product makes sense.

As AI-generated websites and applications become normal, I think this distinction becomes more valuable.

The skill is no longer only producing code. It is identifying the right problem, evaluating the output and turning it into a useful experience.

What’s next

These five tools are only the beginning.

I want to keep building small utilities around the areas I work with every day: AI, web design, frontend development, SEO, content and digital marketing.

Some may be useful to thousands of people. Others might solve a highly specific problem that only appears twice a month.

Both are valuable if they remove friction from real work.

LNNY Tools

Explore the current collection. I’ll keep adding new experiments and utilities as I build them.

Written by

Web design, AI, SEO and digital experiences.