My Last Real Job

I lost my job in the early 1990s when UWCC, a clinical weight-loss company, went out of business. United Weight Loss had been founded by Ed Saltzman, a serial entrepreneur who put together a strong management team that included Steven Silk as CEO. I reported directly to Steve as the company’s MIS director.

Steve would later become the CEO of Hebrew National, where he helped guide the brand from its roots as a New York kosher butcher into a nationally recognized premium hot dog company—selling quality, tradition, and the famous promise that it “answered to a higher authority.”

At UWCC, Steve’s big dream was to get the company listed on the Inc. 500 as one of America’s fastest-growing private companies.

That never happened.

A few years later, though, life showed its sense of humor. My own company, SYS-CON Media, would make the Inc. 500 list three years in a row.

The timing couldn’t have been worse. The recession hit, UWCC never became profitable, and Ed eventually pulled the plug. Steve and the rest of the management team left—and I left with them.

Suddenly I had no job and no plan.

***

What followed was nearly a year of rejection.

I applied to just about every computer and software development job in the tri-state area and landed only a handful of interviews. On paper, the 1990–1991 recession didn’t last long, but in real life it dragged on.

Banks were failing after the Savings and Loan disaster.

Credit was tight.

Oil prices spiked after Iraq invaded Kuwait.

Even when the recession was officially declared “over,” jobs were still scarce.

It felt like the economy had moved on without telling anyone still looking for work.

After almost nine months of searching, I finally landed a job as a PC systems developer at the French chemical company Rhône-Poulenc in Cranbury, New Jersey.

I was their first—and only—“PC guy,” and it showed.

When I arrived, they actually had to scramble to find me a computer. Until then everything ran on mainframes and minicomputers. PCs weren’t part of the plan, and no one was quite sure how—or if—they were supposed to fit in.

The job was a break.

But it came with a new problem: the commute.

I was living at 46 Holly Street in Jersey City and driving all the way down to Cranbury every day. With rush-hour traffic, it took about ninety minutes each way, which I could live with.

What I couldn’t live with was the toll.

The New Jersey Turnpike cost $2.85 each way, plus gas—about $300 a month I simply didn’t have.

Eventually I found a workaround: entering the Turnpike in Elizabeth, which dropped the toll to just thirty-five cents each way.

Over less than three years at Rhône-Poulenc, I put more than 70,000 miles on my Honda Civic.

Proof that sometimes survival isn’t about speed or comfort.

It’s just about finding a way to keep moving forward.

***

In the early 1990s, when I started working at Rhône-Poulenc, RPG—Report Program Generator—was everywhere.

Their IT department didn’t know anything else.

It quietly ran the day-to-day business: payroll, inventory, manufacturing, accounting. No one made a fuss about it.

Most of the RPG code was written in rigid fixed-format columns, tightly bound to the operating system and database. The code wasn’t pretty, but the systems were incredibly fast and reliable.

Those RPG applications just worked—year after year.

Many of the programs they depended on were already old back then, and they still felt indestructible.

Around this time, I developed and rolled out what was effectively the company’s first sales-force automation system—almost a decade before Marc Benioff would make the idea famous.

I’m no genius.

As a PC systems developer, my favorite tool—the one I really got good at—was Paradox from Borland.

I was so into it that I subscribed to several monthly journals published by a company called the COBB Group, including Paradox Users Journal and Paradox Developer’s Journal.

These weren’t glossy magazines.

They were printed in black-and-white, scientific-journal style—the kind that looked serious enough to scare non-programmers.

The subscription wasn’t cheap either: $119 a year for twelve issues.

A few years later, when I launched my first magazine, PowerBuilder Developer’s Journal, I shamelessly copied that exact format and even reused the same $119 price point.

Why mess with a model that already worked?

***

At the time, Borland Paradox was one of the most popular desktop relational databases around, especially in the late 1980s and early 1990s.

My colleagues at Rhône-Poulenc had never heard of PC tools like this. It was a completely foreign world to them.

Paradox made serious database work feel almost friendly.

It bundled a real relational database with forms, reports, and a built-in programming language all in one place. Non-programmers could design tables and run queries, while developers could use ObjectPAL to build surprisingly sophisticated business applications in record time.

On DOS—and later Windows—Paradox earned a reputation for being fast, stable, and extremely developer-friendly.

It was ideal for departmental systems, internal tools, prototypes, and standalone business applications—the kind that quietly ran companies while nobody paid much attention to them.

In the early 1990s, Paradox went head-to-head with dBASE and Microsoft Access. Technically it was strong, but momentum matters.

Once Microsoft bundled Access with Office and the industry shifted toward client-server databases like Oracle and SQL Server, Paradox slowly faded from the spotlight.

Still, it left a lasting mark.

It proved how powerful desktop databases could be and trained an entire generation of developers—including me—to think in relational terms long before web apps and cloud databases were even imaginable.

***

After Paradox, the next tool that completely grabbed me was PowerBuilder.

In the early 1990s, right around the time I started at Rhône-Poulenc, it felt like the future had finally arrived.

Built by Powersoft and later acquired by Sybase, PowerBuilder made it possible to build serious, corporate-grade Windows applications without drowning in C or low-level code.

I knew I had to get my hands on PowerBuilder.

But it wasn’t cheap.

A typical developer license cost between $3,000 and $5,000 per seat.

I convinced my boss, Randy Hompesch, to buy a copy—and he did.

The problem was that PowerBuilder was so new and so revolutionary that there were no courses, books, or formal training programs.

You learned it by doing.

What made PowerBuilder feel almost magical was the DataWindow.

You could visually design a screen, connect it to a database, and suddenly you had querying, updating, validation, and reporting all working together.

Things that used to take weeks—or sometimes months—could be built in days.

For enterprise developers, it was a revelation.

PowerBuilder had been created by David Litwack, who founded Powersoft and designed the original architecture.

 His breakthrough idea—the DataWindow—dramatically simplified building database-driven applications and set PowerBuilder apart from everything else at the time.

By the time Sybase acquired Powersoft, PowerBuilder was already the dominant client-server development tool of the 1990s.

It quickly became the go-to platform inside large corporate IT departments, especially in finance, insurance, telecom, healthcare, and manufacturing.

If you walked into a serious IT shop in the mid-1990s, odds were good that PowerBuilder was running something mission-critical.

***

DUMB LUCK ENVIRONMENT ON MY KITCHEN COUNTER

While all of this was happening at work, life at home was unraveling.

I took a leave of absence, went to Turkey, completed my three months of basic military service in the mountains, and then returned to New Jersey.

When I came home, the apartment was empty.

Everything was gone.

Clothes from the closets.

Shirts and underwear from the bedroom drawers.

Furniture.

Window treatments.

Everything.

I called the Jersey City police and reported a burglary.

I explained that I had been overseas for three months, had just returned, and found the apartment completely stripped.

Two officers walked through the place in silence.

After a few minutes, one of them looked at me and said:

“Nothing’s been stolen. It looks like your wife left you.”

⸻

I really didn’t want to live alone.

So I invited a friend who was working toward his PhD in physics at Stevens Institute of Technology in Hoboken and offered him one of the two bedrooms.

He moved in.

During our late-night conversations at the apartment, I told him about this new client-server platform called PowerBuilder.

“There’s a huge buzz around it,” I said. “It looks like it’s going to take over.”

I needed to learn PowerBuilder, get a better job, and finally end my miserable commute.

The problem was that there were no educational tools.

No courses.

No books.

No clear way to master it.

This wasn’t something you could just install on a PC and study in isolation like Paradox. You needed a real back-end server to connect to.

And that was the irony.

The one place where I could actually do this—at work—was an environment still coding everything in RPG.

For God’s sake.

***

This was all before the internet, at least as we know it today.

There was AOL, though.

You could connect with a 1200-baud modem—something you could buy at 47th Street Photo in Manhattan—and wait patiently as it screamed its way online.

While poking around, I discovered the Powersoft bulletin board.

It was an official dial-up BBS run by Powersoft in the early 1990s, long before web forums existed.

PowerBuilder developers connected by modem to download patches and sample code, read technical notes, and exchange messages with other users—and occasionally with Powersoft engineers themselves.

It was part technical support, part knowledge base, and part community.

For many developers it was the only practical way to learn PowerBuilder beyond pure trial and error.

Long before Stack Overflow, blogs, or online documentation, that bulletin board was where serious PowerBuilder developers went to survive—and level up.

***

This was the most exciting thing that had happened to me since my wife’s disappearance a couple of months earlier.

I spent several nights standing for hours at my kitchen counter.

The modem cable to the only phone jack in the house wasn’t long enough to put my Macintosh Classic on the table, so I had to stand there while connecting.

I was trying to find a PowerBuilder Developer’s Journal—something like my two favorite magazines, Paradox Users Journal and Paradox Developer’s Journal.

I had subscribed to them for years.

I saved every single issue.

They were the most valuable things on my bookshelf—mostly because I didn’t actually own a bookshelf, but you get the idea.

I posted exactly two questions on the Powersoft bulletin board.

First:

“Are there any PowerBuilder magazines out there?”

A couple of people replied.

No.

Then I asked:

“Is there a PowerBuilder book I can buy?”

Again, a few responses came back.

No.

That was… disappointing.

***

Late one night my friend came downstairs to the kitchen to pour himself some coffee.

“Any luck finding a way to learn PowerBuilder?” he asked.

“No,” I said. “There’s nothing out there. Zip. Zilch. I miss my Paradox journals.”

Then a lightbulb went on.

“Wait,” I said. “Why don’t we start a magazine and publish a book ourselves? People need this. This is the hottest thing in IT since COBOL.”

He stared at me.

“What do you know about PowerBuilder?”

“Absolutely nothing,” I said.

He didn’t hesitate.

“Don’t get my name involved. You’re on your own. I’ve got a few months left to finish my PhD and I don’t want any part of this.”

I tried one last time.

“We’ll be fifty-fifty partners. Just hear me out.”

That didn’t help.

If anything, the idea itself seemed to stress him out.

I finally gave up trying to convince my childhood friend and roommate.

That night, right before bed, I posted my third message on the bulletin board:

“Announcing PowerBuilder Developer’s Journal. Respond to this message to receive your sample copy.”

At that point, I figured the worst thing that could happen was more silence.

***

The next morning I woke up early and drove my battle-scarred 1985 Honda Civic to work.

I convinced my boss, Randy, to let me generate daily sales reports using Paradox on a PC.

He looked at me and said:

“You’re useless with the RPG tasks I give you anyway. Go ahead—make my day.”

That was all the permission I needed.

Paradox was my weapon of choice.

Before lunch I dropped three different flavors of sales reports—by sales manager—on his desk.

His eyes went wide.

“How did you do that?” he asked. “And what is this Paradox thing you’re using?”

I told him Paradox was king.

Fast. Relational. No need to fart around with RPG just to answer a simple question.

I came from the relational world—tables, structure, normalization.

Mainframe guys didn’t even know what that meant yet.

To them, data was something you wrestled with.

To me, it was something you disciplined.

In the relational universe, normalization isn’t optional.

It’s law.

One fact. One place.

Break that rule and the data lies to you.

Follow it, and it tells the truth—even when the system gets big, ugly, and expensive.

***

Here’s what actually happened during my short stint at Rhône-Poulenc.

I singled out Larry, the top salesman in the Animal Nutrition division down in Atlanta.

Larry was different.

While most people in the early ’90s were still allergic to PCs, Larry was inseparable from his Compaq LTE. He carried it like it was part of his body.

Larry was also one of the few salespeople who truly needed data to do his job.

He kept pushing the MIS department for timely sales reports—numbers he could actually use while the week was still alive.

What he got instead was a ritual straight out of the Stone Age.

Once a week, someone printed the reports, stuffed them in an envelope, and mailed them to him.

By the time those reports landed on Larry’s desk, they were already history.

And everyone acted like this was normal.

I sat down with him and asked for his top ten reports—the ones he actually cared about.

I told him I’d deliver the information overnight, online, straight to his Compaq.

And if he wanted paper, he could print it at home.

He looked at me and asked:

“How?”

I said, “Give me a few weeks. You’ll have a sales-force automation system running on your PC.”

He was excited—but skeptical.

Honestly, I couldn’t blame him.

***

In the early ’90s, 9,600 baud was considered fast—almost luxurious.

If you were coming from 2,400 baud, jumping to 9,600 felt like the future had arrived.

Screens updated smoothly.

Email didn’t crawl line by line.

You could almost convince yourself the system was real-time.

I installed a rack of ten modems in my office.

On paper, that was more than enough for a hundred salesmen to dial in and pull their reports.

We built and rolled out my Paradox-based sales-force automation system straight to Larry’s Compaq.

And while I was at it, I gave him email too.

Eudora.

Larry, being Larry, couldn’t keep it to himself.

He started showing off his live sales data to other salesmen and casually bragging about his email like it was a new sports car.

Word spread fast.

Before long, Randy—my boss—showed up in my office and asked what I had done for the salesman in Atlanta.

I said:

“Larry gets all his sales reports online. He dials in, updates his numbers daily, and has everything he needs to actually run his business.”

Randy had heard of email.

But in his mind email meant AOL.

I gave him his own company email address anyway.

That’s when things started to get interesting.

***

He went and talked to his boss, Jim Purcell, about whatever strange operation seemed to be running out of my desk.

I gave Jim email too.

Jim didn’t even have a PC on his desk at the time.

After that conversation, he ordered one.

What confused everyone in the department wasn’t the technology.

It was the result.

Somehow, a salesman hundreds of miles away had an application on his computer that actually worked and gave him all the information he needed—when he needed it.

That simply wasn’t how things were supposed to happen.

***

That same week, Randy came back to me and said he’d been getting phone calls.

Every salesman in Atlanta wanted whatever Larry had.

I told him, “No problem. We’ll roll it out to all of them.”

Then I asked, almost as an afterthought:

“Do you want me to order Compaqs for them?”

He said, “Please do.”

About twenty salesmen—all men—got their computers.

Six thousand dollars each.

That was a $120,000 investment.

Which sounded outrageous until you realized the company now had a million-dollar sales system—one the competition didn’t have.

It was a big deal.

And the department was getting credit for it all the way up the chain—from Bernie to Jim to Randy.

They were proud.

***

Remember that water-cooler conversation with my department director, Bernie Kranz, one Saturday morning?

The one where he told me I didn’t fit in the corporate world and should go start my own business?

This was happening right in the middle of all that.

One day, not long after things took off with Randy, I looked up from my desk and saw Bernie standing there.

Just standing.

I knew something was up.

He said he’d been getting phone calls about my sales-force automation system—from other divisions in North America.

They wanted it too.

This wasn’t some small operation.

Rhône-Poulenc was massive—the largest French company, owned by the French government, operating all over the world.

And suddenly whatever had quietly started at my desk was getting attention far beyond our little corner of the building.

Bernie skipped Jim and Randy and talked to me directly.

We rolled out my sales-force automation system to two other divisions in North America.

Guess who showed up at my desk next?

John Wistrich, the president of Specialty Chemicals North America.

“Hey John,” I said.

“How are you doing, Fuat?” he replied.

“We’re rolling out the system,” I told him. “Everyone seems to love it. It’s helping our sales teams tremendously.”

He nodded.

“That’s why I stopped by. I got a call from Paris this morning. The worldwide corporate MIS director is on his way here with his team.”

“They want to see your demo.”

I said, “We’re ready. Whenever they arrive, we’ll set up the presentation.”

John smiled—proud, but a little nervous at the same time—then turned and walked away.

***

The French team arrived.

I presented the system to four executives from Paris, with John Wistrich, Bernie Kranz, Jim Purcell, and Randy Hompesch—my four layers of bosses—all sitting in the room.

We turned down the lights and ran the demo.

When it was over, the senior executive from Paris spoke.

“We will roll out this system worldwide,” he said.

“Starting in Argentina.”

Just like that, my sales-force automation system was set to be used by the largest French company in the world.

Four years before Marc Benioff started Salesforce.com.