1. We gave people access to the most powerful technology that ever existed. They used it to write longer emails.
There’s a simple way to know when technology has been truly transformative. It’s when your work would fall apart if you took it away. Think about electricity, the internet, personal computers — we’re pathetically helpless when they break, and we’d struggle to redesign work without them.
Very few teams have reached this point with AI. Software engineers are the exception — AI has radically changed how they build and ship software. But most enterprise applications of AI don’t stand up to scrutiny.
AI is helping people work a little bit faster — by generating meeting notes, for example, or reviewing long documents — but it’s not transforming how teams work. McKinsey data backs this up — they found that 80% of people say AI has made them individually more productive, but only 37% of organizations report actual business impact.
2. Real transformation means reaching the point of no return.
When I speak to people about AI, I'm most interested in what would happen if you took it away. If any of the following are true, I’m immediately excited:
You'd drop everything to patch the gap
You'd have to hire to replace it
A single day without it would show up in the numbers
Nobody remembers how the work used to get done
We've learned to be grateful for any efficiency AI offers. But AI has the potential to take us much further — to the point of no return, where going back feels unfathomable.
We won't get there if AI only helps individuals improve their own patch of work. The shift has to be operational, not individual.
3. We’ve been chasing greater operational productivity for decades. Now we finally have a chance to catch up.
In 1992, Steve Jobs stood in front of a room of MIT students and tried to explain why the world needed another computer company.
Six years into founding NeXT, he was surprised to learn that the most and least successful companies spent almost exactly the same share of their revenue on IT, around 2%. The key difference was in how they spent it.
The low-performing companies spent their budgets on what he called management productivity applications — tools that helped individuals plan, analyze, and report on their work. The PC. Spreadsheets. Word processors.
The high-performing companies spent their budgets on what he called operational productivity applications — software that changed whole processes, spanned entire teams and ultimately, transformed how work got done.
This data seemed to bother Jobs. PCs and Macs had only ever attacked management productivity. Operational productivity was wide open. For three decades, nobody closed this gap.
Until now.
4. People don’t need another tool or model. They need a complete environment to support enterprise-wide AI.
About a year ago, we set out to design a better version of our original product, Tines Stories. This summer, we launched Tines 3B, still believing it to be a better version of our original product.
Now we know that it’s something else entirely. It’s the answer to two urgent problems:
AI-induced chaos in the enterprise, a problem we like to call wild code
AI’s underwhelming impact on operational productivity
Today, millions of employees are using AI coding assistants to build things, leaving IT and security teams to deal with the fallout. Provided they know these builds exist — and they often don’t — they have to run and maintain these builds while safeguarding against security vulnerabilities, managing AI costs and somehow not burning out their teams, which were never designed to govern software at this scale.
This is exactly the kind of problem we love to solve at Tines. 600+ organizations, from startups to the Fortune 10, already trust us to run their most important workflows. Years of these conversations informed the decisions that led to Tines 3B.
In simple terms, Tines 3B is a complete environment for enterprise AI. Employees can build apps, agents and automations, and organizations can manage everything in one place — connectors and security credentials for proprietary systems and data, AI coding harnesses, observability, deployment, reporting and monitoring.
It's not a coincidence that this mirrors what made AI work for engineers in the first place — they already had version control, CI/CD, and observability waiting for the code AI could write. Every other function has been missing that environment until now.
3B’s real value only became clear to us after we rolled it out internally in July. We realized it reflected a belief we've had since the day we started Tines…
5. The people closest to the problem should build the solution.
Our Director of Finance, Peter Coleman, had known what he wanted to build for years: a way to connect Salesforce and NetSuite so his team no longer had to manually assemble billing data for every deal. Exactly the kind of “operational productivity application” that Jobs described in 1992.
Peter tried to build it himself in our original product, Tines Stories, and couldn't, because he didn't have the technical skill required. Then he made what I'd call the single worst decision in this entire story: he asked me to build it for him.
This process involves sensitive customer data, and the access required was locked down so tightly that I — the CEO — was the only person with permission to touch it. We also didn't have engineering capacity to spare; those precious resources were focused elsewhere. So I built Peter’s idea. And it was bad. I had no idea what a good billing process looked like, no idea what the edge cases were, no idea what actually mattered to Peter's team.
This is one of the key reasons why operational productivity software has eluded us for so long. Until now, the people who understood the problem couldn’t build the software. And the people who could build software didn’t understand the problem.
Peter waited until the technology caught up. On July 6th, we gave every employee at Tines access to Tines 3B, with pre-approved connectors scoped to everybody’s roles. Peter started building that same day. Within a few weeks, it was live. Today, his whole team relies on his billing app. They can’t imagine going back to the way they worked before. And everyone on the finance team is building things that they — and other teams like marketing and senior leadership — rely on.
Elsewhere at Tines, teams have replaced complex vendor-supplied software products, from learning management systems to customer intelligence platforms, with 3B builds. Not because they wanted to save money, although that part is nice, but because these tools couldn’t do the job they needed them to do.
6. Your people already know what they need to build. Let them build it.
I was at an event in Washington, D.C. earlier this year when four-star admiral and former head of U.S. Cyber Command Mike Rogers was asked how he chooses technologies in such uncertain times. His answer has stuck with me: you pick the option that covers the widest possible range of outcomes.
This is what we did with Tines 3B. The complete environment meant we could give people freedom to build — not permission to build anything at all, but confidence that whatever they built would actually make it to production. That's as wide a range of outcomes as anyone can plan for right now.
In times of uncertainty, there's actually very little you can control. But, we figured, we can invest in our people, and we can invest in their ideas.
7. People and ideas are the only real moat. Don’t separate them.
These days, every company has access to more or less the same tools and the same AI models. People and their ideas are the only differentiator.
Some of the best software tools on the market put barriers between people and their ideas. Software has evolved to be opinionated, to do things according to averages, to charge a premium for a product that most customers only use 6% of. Now, suddenly, there’s a compelling alternative — build it yourself in an afternoon.
The statistics from our internal adoption of 3B have been pretty staggering. 90% of Tines employees actively build in 3B. 100% use something that was built in 3B. Around 500 employees have built over 3,000 workflows — many of them experiments — and around 400 of those builds are now relied on to run day-to-day operations. Just 23% include an agentic AI step; the rest run deterministically, which keeps things simple and costs predictable. Dozens of builds serve multiple teams, and dozens more are saving us $50–100k in annual software fees.
For me, the best part is that people no longer have to wait months or years for some external force to change before they can fix their problems. They simply build the fix today. They literally look at their calendar and say, “I have 20 minutes before my next meeting, I bet I can have a working first version live in that time.”
Separate people from their ideas and this phase of AI is going to be really challenging. Give people an environment in which they can build things with AI, and give the organization a way to govern those builds, and suddenly, things become simple. Innovation accelerates. Problems fall away. People become more excited about work than ever. This may sound like an idealized future state. But at Tines, it’s already happening.