I create software for a simple reason. I like to know and learn how things work.
Every child has the same curiosity at the start. They break things. They throw things. They yell in excitement. They learn how the world works by playing with it. I never stopped doing that. I just got better at it.
What makes software unusual is that everything in it is imaginary. You can't touch it, feel it, smell it, or taste it. To talk about it with someone, you both have to imagine the same invisible thing precisely enough that you understand each other. That precision is what got me into writing software. The strange discipline of having to be exact about something nobody can see. And every once in a while, when it's done well, you get to peek into the mind of whoever made it. You can trace the path from the idea through the execution, the repetition, the pain, the giving up, the trying again, the head through the wall, the failure, and then the small private revelation when it works. All for a quiet reward that nobody else will ever quite understand.
That's what got me. That, and games. I hand built every computer I ever owned. Without meaning to, I learned what every part of a computer does, what its symptoms look like when it starts failing, what makes one slow and another fast, and how to set up a local network so my friends and I could play together. That curiosity gave me a foundation I'd never have learned in a classroom.
I ended up working in risk management and e-commerce, building platforms for insurance companies to sell their products online. The word “insurance” makes it sound dull. It's not. That industry compresses everything people now call backend, frontend, UI/UX, devops, and security into a single problem. I did all of it. I worked closely with people at every level of technical literacy, from the deeply technical to the not at all technical. And I learned something that has stayed with me ever since.
We are all just people. We all have a cross to bear and work to do.
My work is to make technology a little closer to the person who isn't as exposed to it as I am. They're tired, stressed, happy, worried, generous, distracted. A whole rainbow of emotion. My job is to remove at least one of the things on their list, so the rest of their day gets a bit easier.
To do that, I had to be close to people. I installed Windows for half the people in my life. I fixed printers, I unbricked routers, I explained, patiently, what a phishing email looks like. I am, visibly, the computer guy. Glasses and long hair. I'm proud of it.
The insurance industry was a great teacher. Most of my work there had very little to do with insurance itself. It had to do with what's around insurance. And the single hardest part of all of it, the part that consumed more of my time than the actual domain ever did, was logging in.
It's astonishing how difficult logging in is.
People don't want to remember passwords. They don't want to manage them. They choose bad ones. They forget good ones. They fall for phishing. I don't blame them. I don't want to remember passwords either.
The mistake every auth system makes is to treat the human as a control mechanism. The human is the weakest part of any control mechanism. So when I started building Workplace, the first thing I built was the identity layer, which I call IdPlace. Use what you already use. If your team signs in with Google, sign in to Workplace with Google. If they have an existing identity provider, IdPlace sits in front of it. If they live in email, send mail from a verified domain. That's enough. The system handles the rest. You shouldn't have to teach your team a new login. They have other things to do.
The second thing I learned, across years of integration work, is that a good database model goes a long way. The shape of the data dictates the tempo of the software's lifetime. I was responsible for integrations across hundreds of teams, which means I spent years in the engine room watching what works and what breaks. The lessons were less technical than people might expect.
The hardest problems weren't algorithms. They were people.
The data businesses need most often lives inside their employees' heads, or their feelings, and forcefully extracting it breaks trust faster than anything else. I watched a lot of harm come from miscommunication between the person who pays for the work and the person who is paid to do it.
That bond breaks for silly reasons more often than for big ones. Sometimes it's compensation. More often it's treatment. Office politics are invisible to the people at the top, so they hire HR departments to manage the human side. Except HR departments have their own politics, and the data that reaches the C level is filtered through several layers of people who'd rather not be the bearer of bad news. By the time a problem gets reported upward, it's been smoothed into something unrecognisable, and the person who left has been quietly recast as someone who needed to leave.
The truth is, you can't tell. You can't tell who your good employee is when performance reviews can be massaged. You can't tell who your best engineer is by counting lines of code. You can't see behind the curtain of data to find the person who silently carries the team. The one who's always there during the bank holidays, who rarely takes sick days, who quietly does the overtime, who doesn't bridge their vacation around the project. That data is enormously valuable. The people generating it are often the ones the upper level of management never hears about, because between them and the C suite there are layers of politics that distort everything.
So companies end up in hiring cycles, often without realising how much money they're losing by not having the right person in the right position. And meanwhile, the competent person who would be that right person has to navigate the hiring market. And the people hiring can't tell how fast that person will become productive, so they hedge and hire someone safer. Everyone loses. It's a cycle, and it can be predicted, and solutions can be offered.
I lost my job over politics. The company I worked for still sells the product I built.
I tried, for a year, to get hired by another company in the EU or the US. I was met with walls of rejections, AI generated emails, and the slowly dawning realisation that the entire hiring and tech industry runs on bad data.
If a company needs a senior developer, good luck finding one. The job ads are written by people who don't know what the words mean. They cram in enough technical jargon to describe an entire team rather than a single person. The person they need will never apply. The people gaming the system will. And they'll provide false data, because there are millions of applicants praying for any opening at all.
The same is true of every job. Waiters, seasonal workers, civil engineers, architects, taxi drivers, florists, factory workers, field staff. We're all swimming in incorrect information about each other. At this point, the way most hiring decisions get made, you might as well crumple a piece of paper and throw it into a crowd and hire whoever it hits.
Insurance, where I'd spent so many years, turned out to be difficult to re-enter if you had no connections. It didn't matter that I knew the entire domain. The first gatekeeper was usually HR, dismissing me for not living in London at the moment I sent the application. It didn't matter that out of 175 tech requirements listed in the job ad, which read like they were hiring a whole department rather than a single developer, I knew 154 of them. One ad asked for Microsoft's HTML5 certification as a prerequisite to talking to anyone who might find my experience useful. To put that in non-technical terms, it's like asking NATO for an official document confirming your autonomous nervous system can blink your eyes.
I had time on my hands. I was unemployed, sick of feeling sad, and I wanted to scratch my own itch before I closed the door on software for good. So I started thinking about Workplace.
Every piece of software I've ever built starts with the same first sentence: we need a list of people who will log in. People are always the starting point. Everything else is built on top.
So I designed Workplace from a different angle than most. I read a lot of business books while I was unemployed. Most of them argue some version of “fake it till you make it.” I disagree with that. It goes against my work ethic. I wanted to build software I actually want to use, and that anyone could use. Not just the company with the largest budget. Why are valuable insights reserved for whoever pays the most? What's wrong with giving a solo founder access to a piece of software that covers 80% of their administrative needs? Do we really need twenty five different SaaS subscriptions just to communicate online and send an invoice?
The angle I designed Workplace from was this.
What if I find the right data model, so I have to write less code?
The work programmers do is mostly covering the bad paths. The happy path is maybe five percent of the work (a number I made up to illustrate); the other ninety five percent is everything that can go wrong. What if I designed the system to have as few bad paths as possible? That's the first angle.
The second angle was harder, because building software is the easy part. The hard part is infrastructure.
I'm an early adopter of cloud computing. Back in 2008, when AWS was new, cloud was genuinely cheaper than running your own server. The marketing hook was different. It was about cost. Today, you need a dedicated team and weeks of back and forth just to spin up a basic service. Compute power that costs thousands of euros a month in the cloud does not exceed the compute power of the smartphone in your pocket. Imagine paying, every month, the price of a new phone, just to keep a cloud server running. That's where the industry is.
So my second angle was hybrid infrastructure. The best of cloud where it makes sense; the best of dedicated physical servers where it doesn't. I run Workplace on dedicated EU hardware that I operate myself. The numbers work out such that I can charge €19 a month and still have margin. Because I'm not paying Silicon Valley prices to serve a customer in Zagreb.
If I can build the software with twenty seven years of experience, I can also build it to run fast, run securely, and reach the solo founder who doesn't need to know any of this. Log in with a passkey. One click, one popup, one more click. Can't be phished. Can't be spoofed. The user doesn't even need to know there's a security model in operation. They just got into Workplace.
While I was building this, I also had to make a living.
We live in capitalism. Workplace was a project, not yet a product, and projects don't pay rent. So I incorporated my own business, started consulting as a B2B contractor, and began taking on clients. This is the part of the story I didn't expect.
For twenty years, I'd worked inside companies where someone else handled the bureaucratic side. I built the product. Other people set up the bank accounts, signed the vendor contracts, sorted the receipts, talked to the accountant, kept the documents in order, handled the compliance forms, dealt with the tax office. I had no idea what was involved. I knew, abstractly, that “running a business” was work. I didn't know what kind of work.
Setting up my own business, across two jurisdictions, taught me. I encountered the other half of business for the first time. The half I'd been quietly shielded from for twenty years.
It was a shock. Not because any individual thing was hard. Each task on its own was small. The shock was the volume, and the fact that none of it ever stopped. Bank papers arriving in the post about an account I'd agreed to without fully reading the terms. Cards expiring at random intervals, the expiry date printed on the card itself, as if anyone reads their own card and commits the date to memory. Vendor contracts. Compliance forms. The signed acknowledgement that I'd been “educated” on GDPR, which was a bureaucratic checkbox more than education. Certificates for fiscalisation. Renewals I didn't know I had until I was late on them. Documents from the accountant. Documents from the bank. Documents from the tax office. Documents I'd received and forgotten and not been able to find when I needed them. A whole pile, growing daily, with no structure.
I bought computer equipment for the work. A server I hand assembled. A desktop with three monitors. A laptop for the evenings. Another laptop because the first one died. A card reader. A mouse. Three keyboards, because I'd seen there were mechanical keyboards and I wanted to try one. (Word of advice: they're loud, and they push my spouse's angry-o-meter to maximum almost instantly.) Each of these came with a receipt, a warranty card, an instruction manual, sometimes a software licence. Each one needed to be filed somewhere I could find it later. None of it was.
I didn't have time to catalogue any of this for my accountant, so I paid my accountant to dig through the pile for me. I still had to get the documents and send them over. And, since the accountant is a human, mistakes happened. Things got missed. Things got categorised in ways I'd need to correct three months later when the tax office sent a polite but firm question.
I reverse calculated, at one point, that I was spending two to four hours a day on this.
Two to four hours. Every day. On things that had nothing to do with the work I was being paid to do.
The pile itself wasn't the worst part. The worst part was that the pile was unstructured. What's an expense? What's an invoice? What's a warranty document? Where is the GDPR thing I signed three years ago? When did the bank send me the terms I agreed to? Email inboxes are glorified piles of unstructured mess if you're someone like me, busy and focused elsewhere. The information existed. It was just unreachable when I needed it.
It felt like assembling Legos from mud, stones and ashes, trying to build something that wouldn't fall apart.
The market response to this problem is twenty five different SaaS subscriptions, each handling one corner of the mess, each charging per seat, each requiring you to learn its own interface. The conversations I had with founders during this period often included sentences where fifteen of the twenty words were product names. We're using Xero for that, Brex for the cards, Notion for docs, Linear for product, Slack for comms, Mercury for banking, Carta for the cap table. I genuinely felt, sometimes, like I was pronouncing a C'thulhu summoning sequence rather than a business conversation.
And there's a smaller, sillier version of the same problem that I want to name because it's exactly the kind of thing the larger software world ignores. To create an invoice in my accounting tool, I had to first create a “partner record” by manually copying the partner's tax ID and address from a PDF on one monitor into a form on the second monitor.
I, a human, was the carrier of data between two screens.
The thing I'm best at, the thing I was being paid to do, is write software that moves data. And here I was, moving it by hand, with my eyes and my fingers, between two computers, because nobody had thought to make those two systems talk to each other.
I'd start the task and think “this will take five minutes.” It always took an hour. I'd be interrupted by an email. I'd lose my place. I'd start over. I'd misread something I'd copied. I'd have to redo it. Meanwhile the actual work waited. I'd lose track of what day of the month it was, because so much of my attention was going to things that had nothing to do with the thing I was best at, which is programming.
This is what I mean by death by a thousand papercuts. Small problems that chip away at the daily work of someone running a business. None of them serious individually. All of them together, an unrelenting drag on attention, money, and the ability to think clearly.
There was a moment when I needed three of my own payslips. The bank wanted them for a loan. I had to email my accountant and ask for them, then wait two days, then receive them, then forward them. Three documents. About my own income. I had to ask someone else for them, and wait for a human to dig them out of a folder.
I sat with that for a while. Why couldn't I just email my-company@workplace.hr with the subject “send me my last three payslips” and get them back in thirty seconds? Why isn't that how this works? We have all this technology. We have computers that can answer almost any factual question in plain language. And the fundamental experience of running a small business is still: dig through your inbox, find the right person, ask them to find the right file, wait.
There's a line I've kept in my head for years, from a game I played decades ago. Every once in a while, a player comes along and reinvents the game. I'm not claiming this for Workplace. I don't think Workplace is reinvention. AI came along recently and reinvented something, depending on who you ask, and even then most people's daily work hasn't changed much from it. What I'm trying to do is smaller and less dramatic. Take the lessons from the early days of the internet, that software should buy you time and not ask for yours, and put them into one coherent product. That's all.
If it works, it works because the lessons were already there. I just stopped ignoring them.
I want Workplace to be a simple thing that asks for minimal input and produces maximal output. You email it the pile of documents you've been meaning to file. It reads what it can, figures out what it can, sorts the receipts from the contracts, extracts the expiry dates from the cards, creates vendor records from the invoices, and presents you with a clean list of what it couldn't figure out, with the original documents one click away. Twenty percent of your time, structured, deliberate. The other eighty percent done by the computer.
Which is what computers are for. To help us, not for us to help them.
One screen, when you sit down to do paperwork, that says: here's what I need your input on this week to keep your business in order. That's it. That's the product.
I don't want to squeeze every last cent out of my customers. I want them to have software that's useful, that isn't intrusive, that doesn't spam them with marketing emails every week, that's there when they need it and works as expected. If you don't log in for two weeks, that's good. It means you were focused on your business, and Workplace was quietly doing its work in the background. Email it the fifty documents you've been meaning to file; check on them when you have time; they'll all be there.
That's what I want. Start with small companies and solo founders. Give them their time back, instead of asking for it. And when they grow, Workplace grows with them. We give you more capacity. We charge you accordingly. We don't punish you for growing. No per seat pricing. No suddenly locked features that you weren't using anyway.
As you grow, Workplace becomes more useful, because patterns start to show. We don't spy on people. But we can tell you who's been there through thick and thin. The colleague who covers for everyone else, the one who'd quietly leave if they got an offer somewhere with better treatment. A small gesture goes a long way for someone like that. We can also tell you when someone is due for a raise.
People tend to quit and join another company specifically to get a raise.
We all live inside the same inflation. Losing an employee and onboarding a new one costs more than giving someone a raise. Workplace can tell you, roughly, how much. €50 a month more, or €18,000 to hire and onboard a replacement with no guarantee they'll be as productive. Which number sounds better now?
Workplace is the combination of what I know about risk management, what I've seen about how people behave, what I learned in the engine rooms of a hundred different businesses, and what I learned in my own back office after I became the one responsible for the back office. It turns simple inputs into useful information. Send an email to your address at workplace.hr. You can run your operations entirely from your inbox if you want to. Mobile UIs aren't really useful if we're being honest. Your inbox already works on your phone, and it's where most of us already live.
I scaled a company from two to a hundred people. I consulted for companies that employed ten thousand. I've seen the operational chaos and the organisational chaos. I lived inside it for two decades. I saw the engine room, the penthouse, and every floor in between. Then I built my own back office and saw what running a small business actually feels like from the inside, with no department to absorb the bureaucracy and no assistant to make the documents make sense.
What I built with Workplace is a telephone line between the floors.
So the engine room can talk to the penthouse without the message being filtered by every floor in between. So the people doing the actual work get heard, and the people making the decisions get the data they're currently flying without. And so the solo founder, the one with no floors yet, has the same operational backbone that takes a hundred-person company a year and three SaaS contracts to assemble.
It's not for one kind of business. It's for any business that has people in it. And paperwork. And the death by a thousand papercuts that comes with both.
That's what Workplace is. That's why I'm building it.
If you've been an employee whose data was treated carelessly, or a founder paying a tax for hiring one person more, or anyone who's ever opened an HR tool and felt like the tool was managing you, you're who I built this for. And if you've ever sat in front of two monitors copying data from one to the other because the software refused to do it for you, you're especially who I built this for. That's where this whole thing started.
The waitlist is open. We launch in July 2026.
— Nikola