A homicide goes cold when everyone has already looked at everything and nobody sees anything new. That is not a problem of effort. It is a problem of seeing. And it turns out the businesses that come to us are usually stuck for the exact same reason.
In the fall of 2022, the Homicide Branch of the Indianapolis Metropolitan Police Department asked me a question that had nothing to do with business. They had a case that had been unsolved for more than two decades. The file was thick. The detectives were good. Everyone who was ever going to confess had not. Could a fresh approach, one built on structured creative problem-solving and this new wave of artificial intelligence, see something the years had buried?
We had done this once before, a first cold case worked with the Simplexity process alone. This time we wanted to pair the method with AI. So we assembled a volunteer crew, nine people with wildly different backgrounds, and we called ourselves Team Monocles. AI experts sat next to people whose relationship with technology stopped at asking Siri for the weather. We met every Monday evening from June through September of 2024. Four months. One self-imposed deadline. Thousands of pages of scanned, handwritten case files.
I want to tell you what happened, because the lesson is not the one you would expect. And because the same lesson is sitting in your organization right now, wearing a different costume.
When you change how you see a problem, you change the problem you see.
Here is what a cold case and a stalled business initiative have in common. In both, the natural instinct is to look harder at the thing you have already been looking at. Read the file again. Run the numbers again. Hold another meeting about the same launch that will not launch. The effort is real and the effort is admirable and the effort mostly does not work, because the problem was never a shortage of effort. It was a frame that quietly went stale.
Detectives are trained to ask "who did this?" That is the right question, and it is also the question that had gone cold. When we ran the case through Simplexity, the eight-step process moved us off that worn groove. Problem finding and fact finding forced us to separate what we actually knew from what everyone had simply been assuming for twenty years. And then problem definition, which is the hardest and most valuable step in the whole method, surfaced a different question underneath the obvious one.
The real problem was not "who did it." The real problem was how do we organize and see thousands of fragments so that a pattern can finally show itself. That reframe changed everything downstream. Suddenly we were not re-reading a file. We were building a system for seeing.
Once the problem was named correctly, the work got concrete. Every team member took a set of individuals from the case and built structured profiles, the same profile template applied to each person, so that comparison became possible. We constructed timelines. We mapped relationships as a social network, nodes and edges, so that indirect connections between people could become visible instead of staying trapped in someone's memory of page 1,400.
We layered AI on top of that human foundation. We ran optical character recognition across the scanned documents to turn handwriting into searchable text. We built a profile generator to draft persons-of-interest summaries automatically. We stood up a social network analysis tool to visualize the web of relationships. We even sketched an application we called the discrepancy catcher, meant to flag contradictions between one statement and another.
Even with a technology as powerful as AI, success still comes down to people and process.
Now the part most case studies leave out. Much of the AI did not work the way we hoped.
The documents were decades old, handwritten, and physically scanned. Our OCR came back roughly 75 percent accurate, which sounds close until you realize that the missing 25 percent is exactly where the ambiguity lives. The discrepancy catcher never got reliable footing on that messy data. Our most ambitious idea, generating "synthetic suspects" we could interview, sat beyond what the tools could honestly do inside a secure, closed environment. We hit dead ends. Repeatedly. We slogged.
And yet. We completed comprehensive profiles for about 90 percent of the key individuals. We turned a mountain of paper into an organized, searchable, structured body of knowledge that had never existed before. We handed IMPD a fresh set of action items, a relationship map, a timeline, and a full set of standard operating procedures for any team that tries this next. The work was solid enough to present to the International Homicide Investigators Association. Captain Roger Spurgeon, the Homicide Branch commander, put it in terms I have never forgotten. He told us what ultimate success would look like, and then, knowing our constraints, he told us what reasonable success would look like. On the reasonable bar, he said, we knocked it out of the park.
That gap between ultimate and reasonable is where almost all real innovation actually lives.
You did not come here for a homicide investigation. You came because something in your organization has been circling the same drain for months and you cannot tell whether the answer is more effort, more budget, or something you have not thought of yet. So let me hand you what the cold case handed us.
The reframe is the whole game. Your team is probably solving the problem it inherited, not the problem you actually have. Before you spend another dollar generating solutions, spend real time defining the problem. The first five answers are usually obvious and wrong because the question was wrong.
Technology sits on top of people and process, never underneath. The AI helped only where a disciplined human method had already organized the ground. Bolt intelligence onto chaos and you get faster chaos. The team's biggest takeaway said it plainly: even with a tool this transformative, it still comes down to people and process.
Prepare the data before you dream about the AI. Our single largest constraint was messy, non-digital input. Most companies rushing to adopt AI have the identical problem in a nicer building. Clean, structured, well-organized information is not the boring prerequisite to the exciting work. It is the work.
Define ultimate success and reasonable success up front. Spurgeon's framing protected the team from despair and from delusion at the same time. Name the moonshot, then name the win you would genuinely be proud of given real constraints. Chase both. Measure against the second.
A diverse, structured team beats a brilliant, siloed one. The AI specialists needed the researchers. The researchers needed the facilitator. The mix was the method. You cannot be your own dentist, and your smartest single expert cannot be your whole innovation function.
When Team Monocles stopped meeting that first Monday in October, something was different, and it was not just the case file. Nine people now knew how to point a disciplined problem-solving process and a stack of new tools at a problem that had beaten smart people for twenty years. That capability did not expire when the project did. It is portable. It is theirs.
That is the whole idea behind how we work. We are not here to hand you a report and disappear. We are here so that when the next stuck problem shows up, and it will, your people already know how to see it differently. You are the hero of this story. We are just the guide who hands you the better map and then steps back to watch you use it.
Let's run your actual challenge through the same process, minus the homicide. Thirty minutes to see whether a reframe is what you have been missing.
Book a call →