I set out to build apps.
Eighteen months ago, Matt and I had a simple idea: build tools that helped people be more intentional with their time. Anti-doomscrolling apps. Productivity tools. Things that made your day better instead of eating it alive.
We weren't thinking about consulting. We were thinking about products.
Four Months in a Tech Stack
What happened next was unplanned — but it shaped everything.
We spent four months deep in the tools. Claude. Claude Code. ChatGPT. Airtable. Supabase. RevenueCat. Xcode. A little Grok mixed in. We weren't dabbling. We were building, breaking things, figuring out what actually worked together and what just looked good in a demo.
That kind of immersion forces you to develop opinions fast. Which model handles what. Where the handoffs should happen. How to build a workflow that doesn't fall apart when something upstream changes.
We weren't learning AI. We were learning how to build with it — which is a different thing entirely.
The Conversation That Changed the Direction
Somewhere in the middle of all that, we started telling people what we were up to.
The reaction surprised us. Not because people were impressed — but because of what they said next.
How do you know all of this?
We didn't have a clean answer. We just knew it because we'd been living in it. But the more we talked to other business owners and operators, the more we realized something: most people weren't anywhere close to where we were — and we didn't even feel ahead.
That's the thing about AI right now. The pace of advancement is so fast that nobody feels current. There's always a new model, a new capability, a new thing you haven't tried yet. So you assume everyone else is further along than you.
They're not.
What Most People Are Actually Doing With AI
Here's what we kept seeing: smart, capable operators using their LLM like an extension of Google. Type a question. Get an answer. Move on.
The more generous version: using it as a content helper. Draft this email. Summarize that document. Generate a caption.
That's not nothing. But it's also not a workflow. It's a habit with no structure behind it — and habits without structure don't compound.
The operators getting real leverage from AI tools aren't using better tools. They're using the same tools with a layer of discipline on top. Context files. Repeatable inputs. Consistent outputs. A system that runs the same way every time so the results are actually usable.
That layer — the process layer — was missing for almost everyone we talked to.
Why Operators Are the Right People to Fix This
Here's what Matt and I bring that most people in this space don't: we've actually run businesses.
Not advised them. Not audited them. Run them — across small, mid, and large operations over two decades. We've sat in the seat where time is the scarcest resource and decisions have real consequences. We know what a broken process costs and what a clean one feels like.
That experience changes how you approach workflow installation. You're not starting from the technology and working backward to fit someone's business. You're starting from how the business actually runs — the real bottlenecks, the real constraints, the places where friction kills momentum — and then deciding where AI fits in.
The tool is the last decision, not the first.
What We're Building at NinetyFive Ai
We still believe in what we started with: helping people be more intentional and effective with their time. The app idea didn't go away. It just pointed us somewhere more immediate.
The operators we talk to don't need another app. They need their existing tools working harder — inside a process that's actually repeatable.
That's what we install. Not a tech product. Not a prompt library. A structured workflow built for how you already work, by people who've worked the same way.