Technology and Background

We are glad you want to learn more about the EASEWORK platform.

On this page you get a look under the hood of the 10. version of the platform that all our solutions are built on. It is intended for anyone who wants to learn more about the underlying technologies, origins, approaches, and visions of the EASEWORK platform.

The current version number is:   10.47.3454

Every single update is tracked by the last part of the version number. This means there have already been 3.454 updates to the platform's code, released in 47 major releases. We are currently on the 10. major release.

But what has led us to this point, and what does the future hold?

I am essentially as old as the internet itself. You could say we were born at the same time. Or at least, the first ordinary people started using the internet as we know it today around 1993.

It all started with a summer course for students about HTML and CSS. My interest was sparked. A year later, I continued with more websites and my first Flash game.

What happens when an introverted, autistic boy finds something exciting and has a lot of free time as a child? He cannot leave his computer and learns everything about the subject.

The big question, however, was how websites work that are not static. A boy called out to me as he was getting off the tram: "Check out PHP, that's what you're looking for." The ride home until I was back at my computer felt endless. And how right he was.

Without any guidance and with an endlessly slow internet connection, which only became faster years later after endless begging of my parents, I taught myself everything I could find and experimented endlessly.

When I had essentially built a social network for my school class just for fun (klassenhp.de) and shared photos from field trips through it, I was asked why I didn't offer it to others as well. A new door had opened. I hadn't even thought of that — I had simply built it out of passion, because I wanted to.

During my exchange year, I spent a lot of time in 2009 programming "onmeet," a social network before Facebook had truly gained traction in Europe. But I realized how much I still didn't know and how unstable and insecure my childhood attempt was.

So I took the advanced computer science course and later went into a dual study program in Business Information Systems with SAP, where I suddenly learned terminology for concepts I had already figured out on my own. From this point on, I could communicate with other developers in a shared language.

Why am I telling this story?

Through my unconventional entry, I have a completely different perspective on software development than what is typically taught at universities or in textbooks.

I have developed my own strategies and systems for creating simple solutions to complex problems, continuously refining my coding — now in the 10th generation. Each time, I threw away the entire codebase and started fresh, because it wasn't good enough yet.

With my experience at SAP as the largest European business software company, I became familiar with the concept of an ERP, as used by large enterprises, and gained experience in various departments, from team-based solution development to consulting, the focus of my dual study program.

From this grew the desire to create a solution that combines business applications with social components, making work easier and more enjoyable. With a clear focus on automation.

I had given SAP a fair chance, but I simply had to pursue my own goals and dreams. I saw no way to get this freedom as a newcomer at SAP. I wasn\'t looking for a job, but a calling. So the only option was to start my own thing right after graduating and register my first business.

Day and night, I worked on a solution for over 2 years and was completely isolated. The most interaction with people in the real world was the hello at the supermarket checkout.

I desperately wanted to find a solution to the problem of how to simplify and automate processes in businesses, and I experimented a lot, producing mountains of notes. Here too, I iterated over and over, taking apart every smallest component of the system.

A simple solution that unites all areas of a business, so that all data is perfectly connected, current, and consistent. Not 1,000 different tools that all work differently and whose interfaces work together more or less well, but a fully integrated system that covers the entire needs of an SME and beyond that can implement specific requirements with ease.

Building on that, a secure and stable layer of automation and customization, so that in just a few clicks, applications can be dynamically adapted live to changing needs and any process in the company can be automated easily — not just by developers, but intuitively by any employee with the appropriate permissions.

Processes are visually designed and usable immediately without any coding. At a level that goes far beyond the simple automations currently available on the market.

It should become possible to set up complex applications in just a few days, instead of taking several years as is often the case today before a system truly works properly. With AI, it may become possible to achieve results in just a few hours. But it requires an absolutely solid foundation so that in the end, everything works smoothly, securely, and at scale.

As a result, I began building my own framework from the ground up, as I was not satisfied enough with any existing technology. The solution had to be uncompromising to create the perfect foundation for my plans with the best possible performance and customizability. A custom development was the only option to achieve my goals. Exactly the solution I envisioned and nothing else.

But after two years of conception without any clients, as I was working on a solution for 10 years from now, my savings had run out and I was forced to bury my dream and work as a freelancer on various projects. One of the lowest points in my life — not being free to work on my passion.

Eventually, I founded a successful startup and helped over 160 restaurants in the first months of the pandemic to launch their first online shop and survive the crisis.

It is very easy to scatter attention when interests are extremely diverse and research is enjoyable.

After creating many different solutions and complete products since the age of 14, starting with the online substitute teacher schedule for my school, homework uploads, or a tool for indexing the yearbook during my exchange year in the USA, or any number of solutions after that, there was a persistent recurring problem:

What good is creating the best solution in the world if nobody knows about it and consequently nobody uses it? The keyword: Marketing!

Starting with affiliate marketing in the hope of earning enough money "passively" to continue working on my platform, I dove deeper and deeper into the world of marketing.

So deep, in fact, that I was suddenly consulting businesses on the topic, and they didn't even know I actually develop software.

Multiple times, I got pulled deeper into marketing, as it was usually the biggest bottleneck in the business, and I supported my clients as best I could there too. Apparently quite successfully, as I was entrusted with budgets of 10,000 EUR per month, among other things, and the ads ran profitably.

But my true passion lies in solving problems, developing systems, and programming software. More analytically crafting the best possible solution rather than throwing a campaign into the black box of algorithms and hoping for the best.

I am very glad to have gained these experiences, as I can integrate these aspects into my products in the best possible way. But I had to completely distance myself from marketing so I could fully focus on my actual core competency (except for my own purposes, of course).

This took two attempts, and no, I will not set up another ad campaign for you or fight with the Facebook Ads Manager! (However, I am happy to provide completely open and honest feedback and suggestions for optimization)

I am building this platform so that your business processes become simpler or fully automated. The tailored solution for your business, to simplify your work (= ease work). I have already "lost" so many years not working on this vision that I now do nothing but pursue it uncompromisingly.

All projects are consistently developed and deployed exclusively on the EASEWORK platform and thus building on the technology presented here. With each project, the platform's capabilities and features grow, benefiting all clients.

The current 10th version and for now final version (meaning it won't be thrown away) of the platform was started in late 2021 and has been developed daily since then. 3.454+ updates have already been poured into it, and we are on the home stretch. We expect to complete all essential components by the end of 2026, after over a decade of development.

There are already 421+ standardized objects in the database to manage all possible business operations and their relationships (as of Jan. 2025).

Each database table has a corresponding PHP class whose instances represent individual entries in the database. This allows the code at every point to access countless metadata and relationships, simplifying development, preventing errors, and automatically generating UIs and forms.

Every single text is dynamically loaded through language files. When the user's language is changed, it automatically reads from the corresponding language file of the other language, provided the field is available. All texts can be translated into any language with one click thanks to AI.

Apps control the actual logic and UI — they define which content is loaded and how it is displayed. They can be nested arbitrarily and called with different parameters.

Themes define the appearance including logos and color scheme, available features, and the URL structure, which can be customized individually per client.

Any number of clients can be managed in parallel on a single instance (tenants, we call them Business). A security layer ensures that when accessing the database, only objects belonging to the user's own Business can be read and modified.

A massive range of standard features, wrappers for API interfaces, and libraries round out the platform, enabling rapid and easy creation of custom solutions.

Even this website with all its features (including the analysis, appointment booking, and newsletter) is built on the platform. You have essentially already become a user of the platform, whether you wanted to or not.

Should an error occur, it is automatically logged and sent to the developers for resolution with all necessary details. Reporting by users is no longer needed.

This is just the tip of the iceberg. If you want to learn more details, you will have no choice but to come work for us. Once 10,000 clients are using the platform, we would be happy to publish a paper on the exact architecture. Until then, it remains an internal secret.

PS: In PHP, we proudly use snake_case everywhere and opening braces on a new line (internal standard), for easy readability and faster typing. In JavaScript and Rust, this is mostly the common style. We view coding as an art form, and the code must not only work well but also be clean and aesthetically pleasing.

From a technology perspective, the platform is remarkably simple in its architecture. A straightforward Nginx web server with a PHP application and a MariaDB database. Add Git for version control and collaboration, Postfix for email delivery, and a backup script, and that is where the complexity ends.

On the frontend, like any website, HTML, CSS, and JavaScript are used. To keep loading times extremely short and responsive, everything is rendered server-side and content is only dynamically swapped via JS.

Through the completely custom development from the ground up and without dependencies on third-party software, the solution is extremely robust, stable, and extremely easy to maintain and debug. Updates can be deployed in seconds and the database updates itself.

Multiple instances are operated in geographically separated data centers, managed centrally and able to communicate with each other. This ensures that in case of potential outages, not all clients are affected simultaneously and backup systems remain accessible.

We have servers in data centers in Germany (Nuremberg, Munich, Dusseldorf, Regensburg), the EU, the USA (e.g. St. Louis), Singapore, and Sydney, Australia. This way, we can always provide services close to the client with low latency to the server. EASEWORK\'s own client data is processed on European servers.

Very few components have been programmed with AI. Mostly when dealing with external interfaces or simple JS functions, where it is faster to use AI than to implement it quickly yourself. The biggest use is for brainstorming, market research, translation into other languages, and marketing.

The fundamental and foundational components were also developed well before the first LLMs became publicly available, and require a stable and secure architecture.

AI is by far not (yet) capable of replacing the competence of an experienced developer who knows their craft. AI is good at repeating what has been programmed a million times before, but novel and innovative solutions cannot emerge from it — at least not at the current state of technology.

In the future, we will run our own local LLMs on our own servers, so we don't have to send our own data or client data to Big Tech. This will enable countless new use cases, as the platform is built from the ground up to optimally support automation.

Already, there are deeply integrated interfaces to OpenAI and Anthropic in the platform, making it easy to implement any client request regarding AI and integrate it into the solution. Whatever you need, we can build it for you. However, data only flows to Big Tech with your explicit consent.

The development has been prepared from A to Z to switch to a completely custom in-memory database solution, which will be implemented in Rust. MariaDB is merely a placeholder until that point.

This will further significantly boost performance and make it easier and faster to process larger data volumes in parallel. Even though the platform is already many times faster than comparable solutions, the actual optimization has not yet begun.

At a later stage, the platform will be operated as a cluster in the cloud, where all instances and users are connected in real time. This means that when data is changed by one user, it will be immediately visible to all other users.

Collaboration across the boundaries of any number of businesses and systems will become possible. We will establish an open standard to easily connect any systems.

A comparable system that achieves this has not been seen in the nearly 10 years since it was conceived and will be an absolutely innovative market first that redefines existing boundaries. Details about this and further developments will be announced via the newsletter.

Currently, placeholder mechanisms are also in use to enable real-time communication for the most important interactions — like in a chat — and instant notifications for any preconfigured events. This is only a tiny preview compared to everything that is still to come with this platform.

We love the innovations in crypto, AI, blockchain, and Web3, especially everything that works completely decentralized, and we have been studying Bitcoin since 2012 when a coin was only a few euros.

Our current focus is on completing the core platform, so we can build solutions in any direction on this foundation. It makes no sense to jump on every hype train.

Since we are completely self-funded, profit maximization is not our highest priority, which allows us to focus on the topics where our passions lie and that help our clients the most.

In any case, concepts are already finding their way into solutions, particularly blockchain technology for signing documents, secure recording of activities, or similar use cases.

It would be theoretically and technically possible to operate the EASEWORK platform completely decentralized as dApps with smart contracts that are connected to each other, enabling entirely new use cases. If the platform becomes "too big," this is a conceivable direction that we are generally open to.

Smart contract-like functionality can also be integrated into existing solutions, although as long as payments are not processed via crypto, this hardly represents a change from the status quo.

The focus, however, as mentioned, is first on completing the platform in its original vision before changing direction or making premature decisions. Regardless of this, we are happy to support projects that build on or integrate these technologies.

Your Andreas

This was probably the longest article you will find on this site. If you have read everything up to this point, I thank you from the bottom of my heart for your interest. Feel free to contact us with any remaining questions.

Are you interested in whether and how we can support you with your next big project?