… This is the fifth in a series of posts about how I ended up where I am today.

My visa was running out. I’d already planned to move back to Ireland before 9/11 happened, and when I returned I moved into a development role. On paper it was a step forward. In practice it was the lowest point of my time at IBM.

Up until then I’d always had some level of autonomy. I could see who used our products and why. I had relationships with developers across the world. I could solution things my own way.

Now I was building software to specification. Given instructions, told to follow them exactly. When I asked about the customers who’d be using what we were building, I was told not to worry about that. On top of that, I was competing with people who’d been working in that architecture for years before I arrived. The whole thing felt like going backwards.

Two things stand out from that period.

My educational background was being used against me again. A senior manager told me directly that I would never be promoted without a university qualification. At one point I spent time coaching a new starter on how to build a presentation. Techniques, structure, how to land the key points. They’d never done one before. After they presented, another manager turned to me and said, “Wow, that’s excellent. Why can’t you do that, Simon?”

I didn’t say anything.

One manager during that time did help me. He took the time to understand where I was and helped me see a way forward. I needed that more than I probably showed.

The Other Side of the Phone

In 2006 I moved to Level 2 support. There’s a misconception about customer support, even inside IBM, that it’s people reading from scripts. At no level is that true, but Level 2 is something else entirely. You’re working with high-value customers or customers who are furious. You regularly have to tell someone they can’t have what they want, because the fix or the feature is just too complex, and they have to come away from that conversation feeling heard and respected.

The people who worked there blew my mind.

In most places I’d been, asking for help was treated as a sign of weakness. Something to be avoided or held against you. In this team, the opposite was true. It was in everyone’s interest to make sure nobody stayed stuck. If your teammate could function without help, it meant your own work got done faster. The whole system only worked if people shared what they knew.

So you’d ask a question. They wouldn’t just give you the answer. They’d coach you to build habits so you could find it yourself. But if the problem went deep enough, they’d give you the answer and then hand you the whole theory behind it. The field of study. The context you’d need to never have to ask again.

There was one rule. You couldn’t ask the same question twice. That was the trade-off. And it was a fair one. It taught you to think about a problem before asking for help, to value someone’s time when they gave it to you. That expectation has stayed with me. To this day it frustrates me when people treat others as their own personal search engine rather than sitting with a problem for a while first.

Knowledge was never used to segregate or to make someone feel small. If you weren’t an expert in something, they made sure you became one. I’d never experienced anything like it. In every other place I’d worked, knowledge was hoarded. People held onto what they knew because it gave them power, made them harder to replace. Here, the opposite was true. Sharing what you knew made the whole team stronger, and that made you stronger too.

Pure Magic

Early on in that role I took a call from an extremely upset customer. Screaming. Threatening. I was in pure panic, completely rattled. I called my manager over.

I can’t do justice to what happened next by describing it in detail, because the skill was in the subtlety. But the customer went from screaming to talking normally, got their fix, and then apologised to me. All from how my manager spoke to them. The words he chose, the pace, the way he acknowledged what they were feeling without giving ground on what was possible.

It was pure magic.

I wanted to understand how he did it. They teach you some of the basics, but I wanted to go deeper. How do you take someone who is at their worst and bring them back? How do you tell someone “no” and have them thank you for it? This wasn’t a trick or a technique. It was a way of thinking about people and what they actually need in a moment of frustration.

Language as a Tool

Those methods became engrained in how people on the team dealt with each other too. Performance reviews weren’t “here’s how you screwed up last year.” They were “here’s how we’re going to make you better in the new year.” The difference sounds small when you write it down. In practice, it changed everything about how people showed up.

It’s about how you use your language. And how you see people’s potential.

Those skills didn’t stay at work. The way you speak to a frustrated customer is the same way you speak to anyone who feels unheard. It saved relationships with customers, yes. But also with family. With friends. I just wish more people understood how much the words you choose and the intent behind them actually matter.

I moved into a team lead role eventually, and two major projects during this time helped shape the next stage of my career. How I became a Master Inventor, and how I built a learning system. Both of those are stories for another time.

The work itself was support for application development, deeply technical stuff that I loved. Towards the end, the focus shifted to supporting an email product. I didn’t want to do that. But my work on the learning system had caught the attention of the Watson team, and that opened a door I hadn’t expected.

Which is where this story picks up in the next post.

---

This is the fifth in a series of posts about how I ended up where I am today. Next: the pivot from conversational AI to agentic systems, and why everything before this was preparation for what came after.

Next Post | Previous Post

… This is the fourth in a series of posts about how I ended up where I am today.

Returning to Ireland, unemployed, I did what anyone without connections does. I mass-mailed my CV to every company in the Yellow Pages. I walked to industrial estates and handed it in at reception desks. Did that for about a month before I got the call from IBM.

Lotus had just been taken over. The interview was unlike anything I’d experienced before. Nobody asked about qualifications. They wanted to know about the technology I’d worked on, the approaches I’d taken, why I’d chosen one way over another. The whole thing felt like a conversation between people who were genuinely curious. It helped that I already knew their products from my Abbey Computer Training days. Lotus 1-2-3 finally paying off.

IBM offered me a part-time contract. I had another offer from a different company at the same time (I honestly can’t remember who), but I went with IBM. Partly because of the technology, but also because they had a wealth of training resources that you could access anytime you wanted. For someone who’d been self-teaching from books and manuals for years, that was like being sent back to the library.

Automating the Boring Parts

My first project was localising Lotus ScreenCam into all the major European languages. Smallest product on the shelf. Myself and one other person were responsible for the lot. The process was semi-manual and slow.

I found it boring, so I automated it. Built tooling to do my work for me, which freed up time to study and tinker with whatever else I could get my hands on. We finished well ahead of schedule.

They moved me onto bigger products after that. SmartSuite, cc:Mail. More of the same work at a larger scale. I got a trip to California to work briefly with the cc:Mail team, which was brilliant.

“How Could You Know That?”

At some point I went for an interview for a full-time permanent role within IBM. They knew my educational background. What they didn’t know was how much of a nerd I actually was.

The main question was to explain, in detail, the entire event chain of interacting with a UI element in Windows 3.1. At the code level. Every message, every callback, every step in the sequence.

I answered it perfectly.

Their response was, “How could you know that?”

I told them a read a book, pulled the system apart, run experiments. The same way I’d learned everything else. Little did I know that reaction would follow me around for years.

They didn’t offer me the permanent role straight away. My contract was running out and I had a full-time job lined up outside IBM. Then the person they’d actually wanted for the role left for another company. I was second in line. They offered me a contract extension.

I told them I had a permanent job waiting for me elsewhere. They changed the offer to a full-time permanent contract. I accepted.

Boston

Some time after this I got put forward for a project that would have meant moving to Singapore. Came second. But the managers involved had something else in mind and offered me a role on a project in Boston instead.

So I went to the US. Again.

The work was localising first-of-a-kind products that IBM was developing. Primarily eSuite, which was a product sadly ahead of its time. Browser-based office applications in the late ’90s. The world and technology wasn’t ready for it, but the engineering was genuinely impressive.

Within the first year I had the localisation work fully automated across multiple systems and operating systems. That freed me up to focus on the first-of-a-kind solutions the management team wanted to build out as products. This was closer to what I actually wanted to be doing.

The World Gets Bigger

My team in Boston was from everywhere. Japan, China, Korea, the US, Ireland. Working that closely with people from different cultures changed how I thought about almost everything. The nuances of communication, the different assumptions people carry, the things that matter in one culture and mean nothing in another. It opened me up in ways I hadn’t expected and was partly why I later visited China, Japan and Korea. I met my wife in Korea.

Of course, the education question followed me there too.

At one team meeting, everyone was asked to share their educational background. When I explained that I’d left school early and never went to university, one of the Japanese team members looked at me and said, “Why are you here?”

It didn’t stop at the meeting. It carried on afterwards until a Japanese exec stepped in and told them to give me the same respect as everyone else on the team.

I still look back on Boston as one of my fondest periods. It was where I started to realise that the world is a much bigger place than I’d understood, and that what seemed so important in Ireland was trivial compared to the work we were doing that reached people everywhere.

The question “why are you here?” was meant to diminish. But over time I started hearing it differently. Less as a challenge and more as something worth actually thinking about. I was there because I kept showing up, kept building, kept learning. The answer was in the work.

---

This is the fourth in a series of posts about how I ended up where I am today. Next: the low point, the unexpected education of customer support, and why how you use your language matters more than what you know.

Next Post | Previous Post

… This is the third in a series of posts about how I ended up where I am today.

The way I got the printing company job tells you something about how things worked back then. After Punch and Programming shut down, some of the customers they’d left stranded reached out to me directly. One of them was a printing company who needed software fixed that I’d originally built in my previous role. I went in, fixed it, and they showed me around.

The place was fascinating. They had everything from state-of-the-art equipment to machines older than I was. And then they showed me a book. Written by a priest, sometime in the 1920s if I remember right, it described in detail how a particular printing machine worked and how it could be constructed. They asked if the machine could be recreated as software.

I said it could. They loaned me the book to do just that.

The Priest’s Machine

My friend Ambrose Clarke and I built it. In fairness, Ambrose did the heavy lifting of recreating the machine’s logic in code. He was a brilliant programmer and that was a serious piece of work. My area was taking his output and converting it into PostScript, the language that professional printing machines understood. The two halves had to talk to each other perfectly or the output would be garbage.

We got paid for the work and at the time we thought it was an insane amount of money. Looking back, it was a pittance for what we delivered. But we didn’t know that yet.

They offered me a full-time job afterwards. I named a salary that felt ambitious to me. They accepted immediately, which should have been a clue. I was getting myself for almost nothing and they knew it.

Hochgeschwindigkeitslaserdrucker

The printing company had a lot of machinery, but two high-speed laser printers stand out.

The first was from the 1970s. You had to flip switches in sequence to boot it up. It stored printing data on magnetic tape. The manuals were in German, and every other sentence seemed to contain the word “Hochgeschwindigkeitslaserdrucker” repeated endlessly. Building anything printable on it was a slow, painful process of switch-flipping and tape-writing.

I figured out a way to get a PC to write the magnetic tapes to the exact specification the hardware expected. You could take a simple text file, run it through my software, and get WYSIWYG mail merge output on this 1970s beast. That one change made the machine genuinely useful to the business.

The second machine was more modern and could handle graphics. Handwriting-style output instead of just print. I built another application for that one too, same idea. Text and images in, professional print-ready output out.

I had a healthy respect for how two of the CEOs ran the business, Pierce and Jean. They were smart operators. But I was building core tools for the company and being paid as though I was an intern.

“You’re Not Qualified for the Job You Have”

When I eventually asked for a pay rise, they refused. Their reasoning was that I wasn’t qualified for the role I was already doing. That I was already doing it didn’t seem to factor in.

My younger self, instead of pushing back, enrolled in a two-year City and Guilds course in C programming and Unix. I passed with distinction. They doubled my salary.

I can’t stress enough that “doubled” still meant very little. The factory workers assumed I was rolling in money because “computer programmer” was a hot job title in the early ’90s. The truth is they were earning three or four times what I was.

But I still didn’t fully understand that. I was learning constantly, working on interesting problems, and assumed that was the deal. You do good work, you get to keep doing it. Nobody sat me down and explained what the market actually paid.

“Show Them the Software I Created”

The moment I can trace the turning point to happened during a customer visit. One of the CEOs was showing some visitors around and said, “Simon, can you please show these gentlemen the software I created.”

I was annoyed enough to say it out loud. “You created it, you can show them.” And I started to walk off.

The customers laughed. The CEO corrected himself and said I’d written it.

After they left he came back to talk to me. He wasn’t angry, and I don’t remember the full conversation. What stuck was one line: “Simon, those guys are in your area and get paid three times what you do, but don’t even know half of what you do.”

I think he meant it as a compliment. I didn’t take it that way.

California, Almost

Some time later, a US company that built software for printing hardware came to visit. They saw what I’d built and were blown away. A few weeks after they left, they offered me a job in California.

I said yes.

Getting a work visa for the US as a non-citizen is time-consuming and expensive for the sponsoring company. They went through the whole process anyway. Months of paperwork proving I had specialised knowledge they couldn’t source locally. Eventually everything was approved. Flights booked. Ready to go.

Then the US company went out of business.

One of the CEOs at my current company asked if I wanted to stay. But we weren’t leaving on good terms by that point. I’d already trained two replacements for my work and had no appetite to go backwards.

Since I had the flights booked and a good relationship with the people I’d met in the States, I went over for a month as a holiday instead. While I was there, another company expressed interest in hiring me. I went for the interview. It was surreal. They asked if I’d ever been to Mexico, then started explaining where I could hop the border to sort out my visa situation. I walked out of that one fairly quickly.

I came home to Ireland, unemployed again. The CEO’s comment should have been a wake-up call. The US offer should have told me something about what I was worth. But honestly, I didn’t connect those dots for a long time. I just knew I needed a job.

---

This is the third in a series of posts about how I ended up where I am today. Next: mass-mailing CVs to every company in the Yellow Pages, and the IBM interview that changed everything.

Next Post | Previous Post

… This is the second in a series of posts about how I ended up where I am today.

After the Motorola interview fell through, I got a second work experience placement at a place called Abbey Computer Training. PCs were just kicking off and everyone wanted to learn word processors, spreadsheets, CPM, DOS. The spreadsheet of choice was Lotus 1-2-3, which would come back to help me later in ways I couldn’t have predicted.

Within a week I was fully up to speed on all the software they taught. By the second week I was pulling things apart to understand how they worked underneath. I was there a few months and genuinely enjoyed it, but it was a tiny operation. The owner and a receptionist. There was no job to offer me, and he was honest about that.

What it did give me was a foothold. Enough experience on a CV to get me through the next door.

The Science of Getting Data Into a Computer

That next door was a company called Punch and Programming. The name tells you the history. They’d started out translating punch cards to computers, and by the time I arrived they’d moved on to keying in all kinds of documents. Pre-OCR, all manual.

Don’t let that fool you though. There is a genuine science behind getting data into a computer fast and accurately. Validation rules, error rates, throughput targets, interface design that minimises mistakes. It sounds mundane until you realise how much of the world ran on people doing this work well.

It was a family-run business and the staff were brilliant. Functional in a way that a lot of workplaces aren’t. There was a lot of manual and technical work, but I got time to tinker and could use the PC in my free time. They trained me on the internals of the IBM PC, from repair work down to how the chips themselves operated. Plus whatever software packages we were selling at the time.

Eventually I got a full-time job out of it. They were paying me next to nothing, but I had no idea what people in my kind of role were actually making. I was just happy to be doing the work.

A Map of Ireland, Taped to a Monitor

Two projects from that period stand out.

The first was for ENFO, the Department of the Environment’s public information service. They wanted an application where users could type in requests for water and air quality reports across Ireland, and it would generate graphs for them to review. A visualisation tool, basically.

The software I built it in has been lost to time. It was a DOS application that used VGA graphics and a kind of no-code step builder for creating applications. Not exactly the tools you’d choose today, but it was what I had.

The tricky part was the map. They wanted a detailed map of Ireland as the main interface, and graphics on PCs were still uncommon enough that nobody had a ready-made solution for this.

So I improvised. I got a sheet of clear plastic, traced over a printed map of Ireland capturing the key regional boundaries, taped the plastic to the monitor, and then sat there recreating it pixel by pixel on screen. It took hours.

When I demo’d the finished application, the room went quiet and then people started asking how I’d managed such a detailed map on a PC. The application got presented to the public for a year or so.

The Leaving Cert Results

The second project carried a bit more weight, in more ways than one.

In 1992 the Department of Education was in the process of computerising the Leaving Cert exam results. OCR was still a young science and their initial tests were far from reliable enough to trust with something this important. If results got corrupted or delayed, every student in the country would feel it.

To make sure everything would be processed on time, they needed a data entry system set up as a fallback. I was the support person for “Key Entry III” at the time, so I spent most of that period in Athlone. Setting up systems, building interfaces, training the UK data entry staff (Irish people weren’t allowed to do the actual keying, for obvious confidentiality reasons), and then validating those results before they went into the Department’s processing systems.

The public only ever see the results landing on a page. Behind that was an incredible amount of work and orchestration to get everything running. The Department of Education staff had mostly all been hired around the same year, so everyone knew everyone. It felt more like a family that happened to be under serious deadline pressure. People worked hard and genuinely enjoyed doing it together.

There was some irony in the fact that I was helping get Leaving Cert results out to the country, having never sat my own. My boss at the time actually told me to never mention this to anyone. Fair enough.

Endings

During all of this, Punch and Programming went through a company change. New name, Data Entry Ireland. New management. The culture shifted. I was made redundant, one of the first to go before the whole place eventually shut down.

I was unemployed again. But some of the customers who’d been left in the lurch by the closure reached out for help, and one of them ended up hiring me.

That next chapter would teach me a very different lesson. The previous years had shown me I could build things that impressed people. What I still hadn’t learned was what those things were worth.

---

This is the second in a series of posts about how I ended up where I am today. Next: a 1920s book from a priest, a German laser printer, and the moment a CEO tried to take credit for my work.

Next Post | Previous post

… This is the first in a series of posts about how I ended up where I am today.

I don’t have a computer science degree. I don’t have any degree. I don’t even have a Leaving Cert. And if you’d met the career advisor who told me to walk out of school a few months before the exams, you’d understand that this was never about choosing an unconventional path. I got nudged off the conventional one and had to figure it out from there.

I want to tell this story because I think it matters. The tech industry is full of people whose paths didn’t follow the script. Most of us spent years not realising what we were actually capable of.

Between Two Countries

I grew up moving back and forth between Ireland and the UK. My parents moved for work, which meant I was always the new kid, always arriving with the wrong accent. In Ireland I sounded English. In England I sounded Irish. This was the tail end of the Troubles, so it went beyond playground stuff. There was a real edge to it.

One side effect of all the moving: I never had to learn Irish. Instead, they’d send me to the library while the rest of the class had Irish lessons. They probably meant it as a holding pen. For me it was a candy store. I read everything I could get my hands on. Looking back, it was probably a form of escape, but it built a habit of figuring things out by reading about them that would carry me further than any curriculum ever did.

School itself never clicked. Most of it felt like going through motions. The exceptions were maths and science, and only because those teachers found ways to make it interesting. The rest I just… didn’t engage with.

“Better Off Not Knowing”

In 6th year we all had the career guidance meeting. The one where someone is supposed to help you figure out what comes next. Mine was short. The advisor looked at my record and told me I’d be better off leaving school than sitting the Leaving Cert. His logic: not knowing my scores was better than failing.

I was seventeen. I took his advice. I left a few months before the exams.

My mother was furious. My parents had divorced by then and I was living with her. I don’t think she was angry at me exactly. More at the situation. At the waste of it. She was right to be.

FAS and the Coin Flip

With no qualification and no plan, I was sent to FAS, Ireland’s national training authority at the time. They assessed me and put me forward for two interviews.

The first was a trades skills course. Woodworking, metalworking, general hands-on stuff with the chance of work experience. The placement was in Darndale. If you want a sense of what Darndale was like in the late ’80s, watch The Commitments and then dial it back only slightly.

The second was a computer science course. Programming, PC technologies, work experience in the field. I’d owned a ZX Spectrum since I was about fifteen. I’d taught myself BASIC, poked briefly around in machine code, spent hours making the thing do things it probably shouldn’t have been able to do. This was the one I wanted.

But the interviews ran in order, and the first course accepted me before the second one could. I ended up in the trades programme. The people were friendly, I made some decent furniture (one of which got stolen) and I tried to make the best of it.

Thirty Minutes

A few months in, they had me sit an aptitude test. The kind that’s supposed to take two hours.

I finished it in under thirty minutes. The examiner was visibly stressed, kept asking if I was sure I was done, if I’d checked my answers. I had.

The next day I got called in. They told me my results were too high for the course I was on. That I shouldn’t be there. They wanted to get me into work experience as quickly as possible and asked what I was interested in.

Computers, I said.

The Pager on the Desk

My first work experience interview was at Motorola. The interviewer was brilliant. We talked about pager technology, programming languages, what the day-to-day would look like. He had a pager on the desk and let me examine it while we talked. I was fascinated. This was exactly the kind of thing I wanted to be doing.

Then he asked where my FAS placement was. I told him Darndale.

He said he’d never heard of it. Excused himself to go ask someone.

When he came back, something had changed. He reached across the desk, took the pager out of my hands, gathered up the other pagers that had been sitting out, and put them all in his desk drawer. Then he started explaining how actually, this job might be a bit too advanced for me.

I was still a kid. I didn’t fully understand what had just happened. It was only later that I pieced it together. He’d gone and asked someone about Darndale, and that was enough. The conversation was over before it ended.

I didn’t get the placement.

What I Took From It

I’d love to say I was angry. That I used it as fuel, that it lit some fire. The truth is I was just confused. I didn’t have the context yet to understand that people would judge you on where you came from before they’d judge you on what you could do. That lesson would keep repeating itself for years.

But something else happened in that period that mattered more. The aptitude test, the Motorola interview, even the trades course. They started to form a pattern I wouldn’t fully recognise until much later. The systems that were supposed to sort people kept putting me in the wrong box. And every time they tested what was actually in the box, the results didn’t match the label.

I didn’t know what to do with that yet. But the next door was about to open, and this time it would be the right one.

---

This is the first in a series of posts about how I ended up where I am today, from leaving school at seventeen to designing agentic AI architectures. Next: how a small computer training company, a data entry job, and a hand-traced map of Ireland set the foundation for everything that followed.

Next Post