The steaks were overcooked, the kids were loud, and Marcus Webb was checking his phone under the table. It is six o’clock on Friday, March twenty-eight, twenty twenty-five, in Fort Lauderdale. Edward Johnson watched his best friend’s thumb moving across the screen and said nothing. Twenty-five years of friendship had taught him when to push and when to wait.
So tell me about the switch, Marcus said, finally pocketing the phone. He turned to Sophia, Edward’s oldest. Pre-med to computer science. Your mom almost had a stroke. She did not. Jennifer’s voice carried from across the table. She called Sarah crying. She called Sarah concerned. There is a difference.
Sophia was eighteen, home from Georgia Tech for spring break. She rolled her eyes. It was not that dramatic. Something clicked in Advanced Placement Computer Science. I was actually making things instead of memorizing what dead people figured out a hundred years ago.
Don’t say that is what I have been telling her. You haven’t been telling her anything. You have been working. Sarah did not look up from helping Isaiah cut his steak. Marcus closed his mouth. Opened it again. Walk me through the hackathon thing. The one you won.
Sophia set down her fork. Forty-eight hours. Four people. We built a functioning app that would have taken a whole semester two years ago. Maybe longer. How. The AI does not write code for you. Everyone thinks that. She paused. You know how Dad describes his first commanding officer. The guy who could look at any problem and ask the one question that made everything click. Edward looked up from his plate. It is like that, Sophia continued. Except it never sleeps and knows every programming language ever invented. But it only works if you know what you are trying to build. She stabbed a piece of steak. The teams that lost. Way better programmers than us. They just kept changing their minds. The AI made them faster at going in circles.
So the bottleneck is clarity, Edward said. Not technical skill. Basically. Marcus’s phone buzzed. He glanced down. Victoria Hartwell, something about initial public offering compliance, and his thumb was typing before he had made a conscious decision. And my professor says by twenty twenty-six, anyone who cannot. Sorry, what. Sophia’s expression flickered. Just for a moment. Then she smoothed it over. I was saying the skills are changing fast. Right. Right. Marcus pocketed the phone. That is really cool, Soph.
The conversation moved on. Basketball season. Science projects. Whether Emma was actually applying to the Naval Academy. But Edward watched Sophia. She had noticed Marcus checking his phone. She had shortened her answer, wrapped it up, made it easy for him to move on. Eighteen years old and already learning to read a room.
At nine thirty PM, the water was black and still. Somewhere across the Intracoastal, a neighbor was playing music too loud for a Friday night. You are still thinking about it, Marcus said, casting his line into the darkness. Aren’t you. I am thinking about eighteen months of clean execution and ringing the bell at the New York Stock Exchange. The reel clicked as he let out more line. After that, we can play with AI tools. Do the whole transformation thing. And if eighteen months is too long. It is not. Marcus turned to face him. Ed, it took us four years to build Axiom. Meridian is still running green screens. Trans-Global just figured out Global Positioning System tracking. Nobody is catching us in eighteen months. What if the rules changed. Rules don’t change that fast.
Edward did not answer. The music across the water switched to something slower. There was this guy Harold, he said finally. At my first job. Defense contractor. Ran the tape library for the mainframes. Knew every tape, every backup, every recovery procedure. The guy you called when everything went sideways. Edward reeled in his line slowly. Then they moved to disk-based backup. Eighteen months. And Harold’s whole career became irrelevant. They gave him a desk in compliance. Make-work stuff. He had gone from the guy everyone needed to the guy nobody called. That is depressing as hell. Essential to ornamental. In eighteen months. You are not going to become Harold. How do you know. Because you are paranoid enough to worry about it. Marcus grinned. Besides, I won’t care if I am irrelevant in twenty twenty-eight. I will be rich.
Edward looked at him. Really looked. What. Nothing. You are doing that thing where you think something and don’t say it. I am not. You are. Twenty-five years. I know the face. Edward cast his line. Watched it arc into the darkness. I just wonder if the initial public offering is the finish line or the starting gun. It is thirteen million dollars. Marcus checked his phone again. That is a finish line. He stood, brushing off his jeans. Look, I gotta deal with this. Board stuff. Go. Marcus headed back toward the house, phone already at his ear. Edward stayed on the dock, watching the lights on the water. I won’t care if I am irrelevant in twenty twenty-eight. I will be rich. Maybe he was right. Maybe the money would be enough. The music across the water had stopped.
After the initial public offering, between April and May of twenty twenty-five, fifteen months to the bell had become Marcus Webb’s mantra. It is seven forty-five AM on Tuesday, April eight, at Axiom Headquarters in Fort Lauderdale. Every decision, every meeting, every strategic choice filtered through that single constraint. Fifteen months until Axiom went public. Fifteen months until he rang the opening bell at the New York Stock Exchange. Fifteen months until the four years of sacrifice finally paid off.
Marcus had grown up on a cattle ranch outside Tulsa. His father was still the foreman there, sixty-four years old, up at four AM, working cattle in weather that would kill most men. Marcus had learned to ride before he could read. Learned that delayed work meant dead cattle. Learned that you solved problems with your hands and your back and your stubborn refusal to quit. He had been a cowboy. A real one. Wrestling had been his ticket out. State champion at one hundred seventy-one pounds, good enough for the Coast Guard Academy. His father had driven him to New London in a truck held together by rust and prayer, shook his hand at the gate, and said three words. Make it count.
Marcus had made it count. He met Sarah at a dive bar. By the time he graduated, they were inseparable. They got married in June two thousand three. The Coast Guard sent them to Kodiak, Alaska. Marcus worked search and rescue coordination, but the real education happened after hours. He brought a laptop and taught himself to code in the long Alaskan winters. By the time they left Kodiak, he had a portfolio of working software and a certainty that he wanted to build things for a living. His last three years of service were in San Francisco. When his five-year commitment ended in two thousand eight, he took a job at a startup.
The next twelve years were a blur of startup chaos. Each time Marcus told Sarah this was the one. Each time he was wrong. By twenty eighteen, he had spent ten years betting on other people’s visions. Six startups. Six failures. Each time he had taken less money for more equity. Soren recruited him with stability. Senior Software Development Engineer. The money was real. He hated it. Two years in the Valley, watching his restricted stock units vest while the commute ate his soul. The cowboy who had grown up solving problems with his hands had become a cog in a machine. Then Sarah got the call to run a charitable trust in Florida. One hundred twenty thousand dollars a year, two days a week. In summer twenty twenty, they sold the Bay Area house and moved to Coral Ridge.
Marcus went to a local startup meetup and met Kyle Mercer and Darren Choi. They had a Hypertext Preprocessor app and a dream. Marcus was ready to bet on himself again. When he walked into Axiom in early twenty twenty-one, the company had two customers and thirty thousand dollars in annual recurring revenue. The tech stack was running on a free tier. Sarah thought he was crazy. But Marcus saw the best product in a forty billion dollar market. They offered him twelve percent of the company. Twelve percent of nothing.
The problem was money. They needed a development team. Kyle and Darren were tapped out. Venture Capitalists would not fund a company with no revenue. So Marcus took out an equity loan against the new house. The cash reserves went first. Then the fifty thousand dollar equity loan. Then the retirement accounts. His four zero one k from Soren, cashed out with penalties. Sarah’s modest individual retirement account. All of it poured into Axiom during two years he took no salary. Everything they had built was gone, converted into shares of a company that barely existed. That money hired the first five engineers. Derek Huang and Prakash Venkataraman. Those engineers built the features that landed the enterprise customers.
Four years later, Axiom had four hundred engineers, two hundred forty million dollars in annual recurring revenue, and a billion-dollar valuation. But Marcus’s twelve percent had not stayed twelve percent. Dilution crushed him. Now he held one point four percent of a company valued at nine hundred fifty million dollars. Thirteen million dollars on paper. Sometimes late at night, Marcus did the math. If he had kept that original twelve percent at nine hundred fifty million. That was one hundred fourteen million dollars. Instead, thirteen million. Still life-changing, but a fraction of what might have been. Big money had become small money, one funding round at a time.
He still drove a twelve-year-old minivan. His sons’ college funds were empty, raided in year two when the credit card juggling ran out. They still owed fifty thousand dollars on the equity loan. He had mortgaged his family’s future on this bet. All of it riding on July sixth, twenty twenty-six. If the initial public offering failed, if the stock cratered, if thirteen million became three million, he would have risked everything for nothing. That was why he could not afford to take risks now. Not with the bell fifteen months away.
After the initial public offering, we can experiment, he told himself. After the initial public offering, we can take risks. After the initial public offering, this all becomes worth it. It was a mantra. A way to defer every hard question until the finish line was behind him. First, he had to get there. The leadership sync started at eight o'clock. Catherine Bell sat at the head of the table. She ran meetings with the efficiency of a metronome. She had been brought in to fix what the founders could not. She spoke the language of board members. Annual recurring revenue growth, gross margins, customer acquisition cost payback periods. Technology was a black box that produced metrics.
Derek Huang, Vice President of Engineering, cleared his throat. AI development tools, he said. GitHub Copilot, Claude, GPT four for code generation. The team keeps asking. What is our position. Prakash Venkataraman, the Principal Architect, leaned forward. Marcus, the developers are going to use these tools whether we sanction them or not. I would rather we get ahead of it. Build an AI strategy. Catherine cut in. Is not having AI coding tools going to directly cause us to lose customers in the next fifteen months. Can you prove that. Derek hesitated. No authorized AI development tools, Catherine said. None. Anyone caught using unauthorized AI tools will be terminated. Immediately. And they will forfeit their unvested equity.
The room went still. We are fifteen months from the bell, Catherine continued. I will not let experimental technology introduce variance into our metrics. She turned to Marcus. Anything to add. Marcus leaned forward. Look, I get it. But we are not some legacy company trying to catch up. We are the transformation. Meridian is still running green screens. Things don’t evolve quickly in the enterprise. We don’t need to chase the next shiny thing. We are the shiny thing. Derek asked if they just wait. We execute, Marcus said. We ship. We go public. The decision was made. Forfeit their unvested equity. That would keep the engineers in line.
The board meeting at the end of April was supposed to be routine. Axiom’s board was dominated by Blackwood Capital. Victoria Hartwell, the lead independent director, pulled up a report. More enterprises are using AI to build software, she said. McKinsey says it is accelerating. We are not in the conversation. What is our response. Marcus had prepared for this. Victoria, we already have AI, he said. Predictive features. Demand forecasting. What you are reading about is using AI to build software. Why would I introduce experimental tools right before an initial public offering. That is how you ship bugs and spook the underwriters. We execute the plan. After the initial public offering, we can have this conversation properly.
Later, Kwame Okonkwo, a new board member, caught Marcus in the hallway. I’ve watched companies do this, he said. Always after something. We are not those companies, Marcus said. He went back to his office. Meridian Freight had gone dark. No product announcements. No conference presence. Their stock chart was a flat line at nine dollars. The analyst commentary was brutal. Meridian continues to serve its existing customer base but shows no signs of innovation. Marcus closed the laptop, satisfied. Meridian was exactly what he thought. A legacy company coasting on existing contracts. He would check again after the initial public offering.
At home, Sarah told him he missed his son’s basketball game. Fourteen months, Marcus said. Fourteen months and all of this becomes worth it. MJ has stopped expecting you to show up, Sarah said. That is what worries me. Late at night, Marcus sat at his desk. He typed a heading. Where does work get stuck. He tracked the last three major features. The average was fourteen weeks. Fourteen weeks from we should build this to customers can use it. Requirements gathering took two weeks. Waiting for approvals took two more. Development was only three weeks on average. Everything else was waiting. Queues. Checkpoints. Delays. He could fix this. But Catherine would ask questions. Any significant process change would introduce variance. Variance spooked investors. He highlighted the analysis and pressed delete. It is not looking away, he told himself. It is prioritizing.
Edward sits at the edge of the dock at five thirty AM on May first, twenty twenty-five. He has been coming here every morning since spring break. He is wrestling with Marcus’s words. I won’t care if I am irrelevant in twenty twenty-eight. I will be rich. Edward cannot stop thinking about Harold, the expert who became obsolete overnight. Caution had kept Edward alive in the Persian Gulf, but caution could curdle into paralysis. He is forty-three years old. If he doesn’t figure this out, the board will bring in someone who will.
Jennifer had given up her career to raise their three daughters. Everything rode on Edward’s salary and his relevance. But there was something more immediate. The medication tracking feature. It had been in the backlog for eighteen months. It could save thousands of lives by flagging drug interactions. It sat there, stuck, because the engineering team was buried in maintenance. The process was designed to produce activity, not outcomes.
Rachel Torres, his chief of staff, told him engineers were using AI tools on their own. Copying code into ChatGPT. Using personal Copilot accounts. If they are doing this despite the risk, Edward said, there is a need we are not meeting. Figure out why they are in pain. He brought in consultants from Stratton McKelvey. They proposed a transformation roadmap. Twenty-four to forty-eight months. Ninety days just to assess. Edward asked if any of them actually built software with AI tools. Silence. We are consultants, the lead said. Edward fired them. Saved two million dollars and two years of nothing. They can’t teach what they have never learned, he told Rachel.
He tried the vendors next. Slick slides. Impressive demos. But none of them could help him ship the medication tracking feature. He would have to figure this out on his own. At two AM, he almost called the consultants back because he was scared. Jennifer told him to delete the email. You know how to build, Ed. Trust that. On May twenty-four, he downloaded Claude Code, Copilot, and Cursor. He started small. A Python script to parse log files. It took a junior engineer two days. Edward had it working in three hours. It was elegant. It was fun. He felt the joy of building come back after eight years away.
Sophia came home for spring break and found him on the dock. She walked him through her workflow. Prompt patterns. Context windows. My professor says don't fall in love with the hammer, she said. Fall in love with the house you are building. Edward realized the transformation industry was selling expertise it did not have. The real expertise was in dorm rooms and on docks. He flew to Atlanta to speak to her computer club. He didn't lecture. He asked them what they were building. He learned more from those twenty-year-olds in two hours than from every consultant he had met.
Back at Riverton, he gathered six engineers. Where does work get stuck, he asked. They mapped the process. Not the official documentation, the reality. Requirements to design took three weeks. Waiting for environment provisioning took two weeks. Writing manual test scripts took two weeks. HIPAA review took three. Thirteen weeks total to ship a single feature. Eleven weeks of waiting. Two weeks of work. Eighty-five percent of cycle time was queue time. This is not a technology problem, Edward said. This is a process problem.
He went to the Chief Executive Officer, David Aldridge. Give me ninety days and one feature, he said. The medication tracking feature. If we can ship it in ninety days, I have proven the approach works. David agreed. Edward dissolved the existing governance committees and created the AI Software Development Life Cycle Board. One board. Decisions in twenty-four hours or less. No more waiting for meetings. He reassigned twenty Quality Assurance engineers who were doing manual testing. They knew the system better than anyone. Instead of clicking screens, they started building features.
The Infrastructure team pushed too fast and the staging environment went down for twelve hours. The board wanted to shut everything down. Edward told them the approach didn't prevent the failure, it made it visible and learnable. On day eighty-seven, the medication tracking feature passed final testing. They had cut the cycle from thirteen weeks to under five. AI handled three of the bottlenecks. The other four were just process bloat. The feature that could save lives was finally going to ship. Not because of a transformation roadmap, but because they asked a simple question. What is stopping us. And then they fixed it.
Thirteen months to the bell, Catherine Bell sat at the head of the table at Axiom in June twenty twenty-five. The elevator pitch was simple. Axiom Disrupts freight management. Twelve percent market share. Our competitors are years behind. Marcus handled the AI questions. No material changes to development methodology are planned, he said. We protect the registration. Prakash Venkataraman told him Meridian had gone dark. No product announcements. They are dying, Marcus said. We have a four-year head start.
In Manhattan, Margaret Sullivan at Hartley Shipping told him the roadmap was slipping. AI-powered route optimization was promised for twenty twenty-four. Now it is twenty twenty-five. My team jokes you have gone business casual, she said. If Meridian shipped the features you keep promising, I would have to take a look. Marcus laughed it off. He caught Derek Huang at his car. Are we moving fast enough, he asked. No, Derek said. We are protecting a trajectory instead of building a future. I would map the process and find where work gets stuck. Marcus stood in the empty garage. He could fix this. But the bell was twelve months away. After the initial public offering, he told himself.
In October twenty twenty-five, the second wave of pilot selection began at Riverton. Edward used the excitement around AI to force middle management to fix the process. If a team did not have four hours to map their value stream, he didn't let them in the pilot. Then a crisis hit. A nurse at Saint Augustine Memorial accessed the system during a code blue. The medication module locked up for eleven seconds. The patient was fine, but the near-miss was filed. The root cause was automated logging that wasn't optimized for real-time access. We fixed one bottleneck and created another, Edward said. He paused the rollout and added a stress testing requirement. He owned the mistake publicly.
Johnny Morrison, the transformation lead, went to the board to shut Edward down. A patient could have been harmed, he said. Edward told the board that under the old governance, they had three incidents last year and no one knew why. This time, they found the root cause in hours. The board let him continue, but Johnny joined the AI board. Friction became accountability. The Security team cut their response time from fourteen days to three. They removed a manual compliance gate and then hit a HIPAA audit finding. They had assumed a gate was waste when it was actually load-bearing. Edward added a new step. Interview the person who performs the gate before you cut it. Learn what it actually does.
Kevin Nakamura, the veteran code reviewer, was ready to quit. He thought AI was going to replace his fifteen years of intuition. Edward bought him coffee. It is a filter, not a replacement, Edward said. The AI handles the syntax so you can focus on the architecture. Kevin became the pilot’s most rigorous auditor. By November, the medication feature was in twenty-three hospitals. Already catching drug interactions. Real patients, safer. Sandra Williams, the Chief Financial Officer, backed the results. She told the board the process improvement was the innovation, not the AI. Johnny Morrison was reassigned under the Chief Technical Officer. He resigned. You can’t stop the flood, Edward thought. You just build channels for the water.
By February twenty twenty-six, the quarterly review at Riverton told a story that numbers rarely tell. Ten percent. That was the organization-wide improvement. The pilot teams were up forty percent, but averaged across four hundred engineers, it was ten. The approach worked when Edward was in the room, but it didn't scale sequentially. The Infrastructure team optimized the wrong thing. The Mobile team shipped features faster that nobody wanted. The framework was powerful enough to be weaponized by managers looking for headcount reductions. Success was supposed to bring better problems. Instead, it brought chaos.
Edward realized he was the bottleneck. He had built a dependency, not a transformation. He spent his political capital on the pilots and didn't have enough left to restructure the whole organization. So he took a job at Cascade Technologies. They offered him a blank canvas. A hundred-and-fifty-thousand-person company where he could build a parallel organization from scratch. No legacy process to fight. He wasn't abandoning the work. He was going where it could actually move.
At Axiom, Marcus was oblivious to the shift. Whitfield Capital upgraded them to a Strong Buy with a twenty-two dollar price target. I was right, Marcus thought. All the doubters were wrong. Then customer success flagged twelve accounts as at risk. Customers were asking about Project Prometheus. A mysterious startup was approaching them with free migration and features Axiom had not even scoped. Marcus told them to stay warm. We are two months from the bell, he said. Once we have public capital, we can accelerate.
On July fifth, twenty twenty-six, the news alert changed everything. Meridian Freight was behind Project Prometheus. They had rebuilt their entire tech stack in fourteen months using a tiger team of fourteen engineers and aggressive AI tools. They had kept it secret until the day before Axiom’s initial public offering. Fourteen months. Fourteen engineers. Marcus’s four-year moat was gone. The silence had not been Meridian dying. It was them building. The migration offer included a ninety-day switchover guarantee. They were specifically targeting his customers.
Marcus Webb rang the opening bell on July sixth. The sound was resonant and deep. But the stock did not pop. It opened at sixteen dollars and started to slide. fifteen point eighty. fifteen point fifty. By market close, twelve dollars. Twenty-five percent down in one day. Marcus’s stake was worth ten million dollars on paper. But he was in a one-hundred-eighty-day lockup. He could only watch as the value evaporated. He found out a senior architect had already left for Meridian. Fourteen people beat four hundred. Fourteen people with the right tools and the right questions.
In August, the stock hit ten dollars. Catherine Bell resigned. I build companies for exits, she said. I don't rebuild them after exits go wrong. In September, it hit eight dollars. In October, seven. Marcus finally called Edward. Tell me how to catch up, he said. Edward told him there was no playbook. Start with the outcome. Map the process. Separate real constraints from invented ones. Prove it small. Marcus tried to learn to build with AI at three AM. He failed. The fifteen-year gap was too wide. He hired consultants for two point four million dollars. They proposed an eighteen-month roadmap. They never once asked what was stopping him from shipping.
By November, the stock was at six dollars. The board offered Marcus garden leave. Twelve months of paid silence. He accepted, but with one condition. He wanted access to all the technical training materials. He would take twenty twenty-seven to become relevant in twenty twenty-eight. He met Edward on the dock on Christmas Eve. He showed him a compass Edward had given him. Valuable to someone willing to admit they are lost, Edward had said. Marcus was lost, but he was finally asking questions.
A year later, in December twenty twenty-seven, the work continues. Marcus’s first client, Plainsight Field Services, had a feature stuck for eleven months. He mapped the process and found forty-three days of waiting for five days of work. He fixed the decision path. The feature shipped in three weeks. Sophia got a job at Cascade. Edward’s parallel organization is growing, but it is starting to develop its own politics. The looking never stops. The process you never examine becomes the bottleneck you cannot fix. Success can turn into a new kind of blindness if you let it. Marcus still carries the compass. He is finally asking how to keep looking.