How CineLog Began — A Founder Story
The story before the product. Two founders, a partnership, and the months between an Instagram DM and the decision that this was real.
CineLog didn’t start with a screenplay parser or a shot list or any of the things later articles in this series are about. It started with a message.
This article is in my voice — Ivan, one of the two founders. Olexiy is the other.
February 7, 2025
I’d been running a project on the photography side of the creative industry. It was real enough to keep me thinking about tools for visual professionals, but I’d been quietly orbiting around film for months. The more I looked, the more obvious it was that the people I most wanted to build for weren’t photographers — they were filmmakers. The day-to-day of a film production was full of the kind of problems software could genuinely help with. There were existing software solutions, of course. Most of what I tried felt clunky — minimum-effort products built to make money off filmmakers, not to serve them. I wanted to make something different: built with intention, with care, for the people in the industry.
So I started watching how directors actually work. I found Olexiy on Instagram — a working director here in Austin, building out his career, putting together an online course teaching filmmakers what he’d learned. His posts had a particular quality I responded to: focused, generous with what he knew, honest about the parts that were hard. That last part mattered. People who tell you the truth about the work are usually the ones worth working with.
I sent him a DM. We met on February 7, 2025.
Two hours, no friction
What we both noticed in that first conversation was how quickly we stopped explaining things to each other. You learn a lot about a potential collaborator by how much context you have to set up before you can talk about the real thing. With Olexiy, that overhead was zero. He described the parts of pre-production that drained him. I described the parts of building tools that felt missing in the existing software. We kept landing on the same observations from different sides — me from the builder’s chair, him from the director’s.
The conversation didn’t end with “let’s start a company.” It ended with “let’s see if there’s a project here.” That’s a more honest place to start. Most ideas don’t survive the next four weeks of actually thinking about them. If ours did, we’d talk again.
February through May: from notebooks to a working PoC
We didn’t quit anything. Neither of us could afford to. Olexiy had his directing work and his course. I had a day job and a side-project of my own already running. We weren’t going to pretend any of that was about to change.
What we did was set up a standing meeting: Friday evenings, every week, blocked on the calendar. We talked on the phone and online in between, but the Friday session was the dedicated time — and having a dedicated time turned out to matter more than we expected. A weekly slot you don’t let slip is how you keep momentum through months when there’s nothing user-visible to point at.
The first few sessions were almost entirely planning. We sketched feature lists. We talked through what a real shot list app would feel like — not just the features, but the rhythm of using it. The shot list was the heart of the product. The script side, at that point, was a supporting panel: import a screenplay and display it next to the shot list so a director could read along while breaking it down. The full-blown screenplay editor we have today only came later that summer, after we ran into Fountain syntax and realized the script side deserved much more than a static display.
We walked through a director’s pre-production week line-by-line: which parts could be made easier, which parts were unavoidable, which parts nobody had even tried to make easier yet. Olexiy brought the texture of how the work actually happens. I brought the question of what was buildable in what amount of time.
That’s the part of starting a company nobody writes about, because there’s nothing to show. No commits to a repo. No product. No demo. Two people in Austin, every Friday evening, with notebooks and a growing list of “what if we built this part right.”
Within a few weeks of that I was writing code. Not the full product — the load-bearing pieces we’d need first. The shot list as the core surface. A screenplay parser that could read a script and display it next to the shot list. A basic scene model that connected the two. The early code was narrow and far from finished, but it gave the Friday discussions something concrete to react to. Showing Olexiy a working scene parser and watching him try to break it surfaced more product decisions than any whiteboard session would have.
By the end of May we had a proof of concept. It was small and fragile — it crashed regularly, and there were a lot of edges that weren’t smoothed yet. We weren’t ready to show it outside the room. But it covered the spine of the workflow — import a screenplay, see it parsed, generate a starting shot list from it — and that was the threshold we’d been working toward.
What kept it going through those months was that the conversations didn’t get worse. Most “let’s start something” conversations decay — the first week is exciting, the third week is half the energy, by week six you’ve quietly moved on. Ours didn’t decay. Every meeting added more clarity to what we wanted to build, not less. That was the signal we were both watching for, even if we didn’t say it out loud.
Early April: shit got real
There’s a moment in any project where it stops being “we should do this” and starts being “we are doing this.” For us, the first one came in early April 2025, when we picked a name.
The name was good. Short, distinct, easy to say, easy to look up. The catch was the .com domain — it was on the aftermarket, priced like an actual asset, not a side-project flier. We were two months in, with code that barely worked, no entity, and no revenue. Spending that kind of money on something still this early was a big ask.
We sat with it for a few weeks, then bought it on April 25. That call turned out to be one of the most important decisions we made in the entire pre-product stage. A solid .com still carries authority on the open internet in a way that’s hard to overstate. The .app, .io, .dev variants — they’re fine, but they read as something that might disappear in six months. A .com says you’re committing to this for the long run. Every conversation we’ve had since with a customer or a partner has been quietly easier because of that decision.
The name and the domain were the first public-facing artifacts of the project. Everything else — entity, trademark, the finished product — was still ahead.
June: the commitment moment
The PoC sat in front of us through early June. It worked enough to look at the thing we’d been talking about for four months and recognize it. The shape was right. The rhythms were right. You could pick it up and immediately see what the finished version would feel like, even though what was in front of us was three feet short of finished.
That’s the moment that decides whether you keep going or quietly let the project drift back onto the side-projects pile. We sat with it for a few days and asked each other the only question that actually matters: do we believe in this enough to take it seriously?
We did. That was the commitment moment.
What “commit” actually meant for us
Commit didn’t mean we quit our jobs. It meant we restructured how we spent our hours. It meant we agreed that this was the project — the one that got the protected time, the one our other work would make room for. It meant we set ourselves a rough schedule and started shipping the thing instead of just talking about it.
It also meant a paper trail. On June 8 we filed the LLC with the Texas Secretary of State. Three weeks later, on June 27, we filed a trademark application — which, at the time, felt like belt-and-suspenders paperwork, but which we’d later come to understand was existential for the brand. The full story of why is its own article, somewhere further down the series. The short version: if you’re building something with a name you care about, file the trademark sooner than your gut tells you to. Done early, it costs a few hundred dollars and an afternoon of forms. Done late — or not at all — it can cost you the name itself, and everything you’ve built on top of it.
Building a real product while keeping the rest of your life intact is a particular kind of effort. You don’t get to put everything else on hold. The day job is still there. The personal life is still there. The thing you’re building has to find its hours inside that, not on top of it. The trade-off is that the work goes slower than it would in a “we both quit and went full-time” sprint. The upside is that you don’t burn out at week eleven, and you can keep the thing alive long enough to actually find out whether it’s any good.
The part nobody says enough
The other thing we learned in that first stretch — and I don’t think this gets said enough — is that the experience of building something that matters is dramatically different when you’re doing it with someone.
The progress is the same on paper. The feeling is not.
When you’ve spent a Saturday afternoon making something a little better, and there’s someone else who cares about it exactly as much as you do, who you can show it to that night, who’s going to be tired in the same way at the same time — the work doesn’t drain you the way solo work does. Partnership isn’t just division of labor. It’s the multiplication of how much of yourself you’re willing to put in, because you’re not putting it in alone.
The encouragement isn’t something you ask for and the other person gives you. It’s structural. It’s just there, because they’re in the same fight.
By the end of June 2025 the foundation pieces were in place: a working PoC, a real commitment, an LLC, a trademark application. Four months of head-down work later — October — the first version went in front of real users: a small alpha group, the first people to try CineLog outside our own room.
The next article picks up at the start of that stretch — the technical work that began the moment we said yes.
Next: Why Parsing a Screenplay Is Harder Than It Looks — the first real technical problem we ran at, and what it taught us about screenplay parsers nobody warns you about.