She Was Working Until 3 a.m. Every Quarter — What She Built Next Should Be a Lesson for Every CEO

Every company has an innovation strategy. Most are wrong; not because the ideas are terrible, but because they often begin in the wrong places. They often start in a boardroom with a pricey consultant’s presentation, rather than with the person in your prospective customer’s organization who is quietly struggling with a broken process that no one cared enough about to understand or repair.

I learned this by speaking with a woman I met at a user conference several years ago. This is in many ways her story — what she told me about her job fundamentally informs my perspective on technology, leadership and where real change often comes from.

The part most never think about

In a large enterprise, close to half of all deals are frequently closed in the very last week of the quarter, many in the final hour(s). Predictably, the sales organization gets the spotlight for landing those deals.

However, behind every signed contract is someone who has to pull information from numerous systems, cross-check with legal, verify with finance and chase compliance threads across dozens of systems. It is complex, high-risk work that is often performed under strict deadlines that couldn’t care less how many threads you’re juggling.

She was one of those people. She processed possibly hundreds of contracts every quarter, and the last two to three weeks in each quarter was working past midnight and waking up early the next day to begin again. She had young kids whom she rarely saw during those periods. The frustrating thing wasn’t that she was bad at her job — in fact, she was very good at her job. It was the process that would wear people down.

Taking initiative

She had heard about AI and intelligent automation — not simple macros but true automation tools that could pull data from other systems, process and analyze that data, and then move forward in a workflow. So without a mandate, without a budget, without asking anyone’s permission, she began to build.

She created what we would now call an AI agent. These “agents” took care of the mechanical aspects of her workflow — pulling documents, processing them for the right data, closing tickets, submitting paperwork and routing for signatures. It changed everything. She was able to focus on complex judgment calls that truly needed her expertise while delegating the busywork. She even gave her creation a name, “Connie,” a play on the word Contract Ops.

The end results were astonishing. She was able to accomplish her work 10 times faster. Not only that, but the quality of her output improved. And, most importantly for her, she was able to spend more time with her family during quarter-end.

Spreading without a mandate

She didn’t pitch this to leadership. She didn’t submit a formal proposal. Yet, in a matter of days, a team of 20 was using her agent.

And of course they were. The solution quickly rolled through the financial department since the people adopting it were dealing with the same pain. And when a contract can’t get signed in the right quarter, revenue slips. For large enterprises, we’re talking millions of dollars. Her solution didn’t just improve her quality of life — it directly protected the business.

The pattern that changed how I lead

That conversation has stayed with me over the years. I’ve seen the same dynamic everywhere I’ve led teams. The best innovations never start with someone thinking, “I’m going to build something that goes viral across my entire company.” They start with one person saying to themselves, “I currently have to do this thing twice every morning, but what if I only had to do it once?“

That’s it. It’s that small.

What makes these kinds of solutions so powerful is that the person building them is the domain expert. They understand the work deeply in a way that no consultant or executive ever could.

It’s not about knowing software or understanding programming; those skills are almost irrelevant, especially today with the tools we have. It’s about knowing your job extremely well. A lawyer who knows contract law. An ops person who understands procurement. A finance analyst who can reconcile in their sleep. When people solve their own friction, the solution fits because they built it for the reality of the work — not for some process map on a whiteboard.

Eventually, they tell a colleague who has the same problem. And it spreads. Not because anyone mandated it. Because it simply makes sense.

What this means for leaders

That woman in contract ops wasn’t just solving her own problem. She was demonstrating something that most companies still haven’t internalized: The people closest to the work are often the ones best positioned to transform it. All they need are the tools that meet them where they are and make sense for them.

This customer-centric idea forms the center of everything we’re building at Sauce Labs. Today, software developers spend huge amounts of time on testing — writing scripts for complex and simple customer journeys, maintaining them as the software or the environment changes, debugging failures — a lot of which creates busywork and takes time away from the software that they actually want to write and ship.

It’s the same pattern that gave birth to “Connie”; talented people who are trapped in repetitive tasks that are critical and hard but can and should be automated so that they can focus their time on judgment and creativity.

So we built AI that does exactly that. Sauce AI for Authoring lets teams describe what they want to test in plain language and get working code back in minutes instead of spending days writing it. Sauce AI for Insights analyzes test data to surface what actually matters, pointing teams to failures and probable causes instead of leaving them to sift through noise. The goal follows the same principle that the woman showed me at that conference many years ago: Mechanize the mechanical, so humans can do the human work.

My job as CEO isn’t to design those solutions from the top. It’s to make sure that there is nothing in the organization that stands in the way of the people who will. That means allowing people to experiment, making it safe to try things that may not work and paying attention to the roles in your organization that are invisible until something breaks.

Because somewhere in your organization right now, there is a person working hours that would appall you, doing work that a machine should be doing, solving it quietly because nobody asked. The question isn’t whether they exist. The question is whether you’ve made it easy enough for their solution to spread — or whether you’re going to wait until they burn out and leave.