One calendar app, one to-do app. I ran that pair for a long time. Appointments in the calendar, tasks in the list. That's how most people do it, and since I'm not someone who switches tools constantly, I figured it was enough.
Then, at some point, it started to feel hollow. Both apps did their jobs, and yet the things worth keeping weren't ending up in either of them.
The to-do app did exactly what its name promised: a place to write down whatever was bearing down on me right now. Write it, do it, clear it. It handled that job well. The problem was that clearing it left nothing behind. I'd spend a month ticking things off, and if you asked me at the end of it what I'd actually done, I had no real answer. The records had existed. Nothing remained.
So I started keeping a diary in my calendar
I began writing the day into Google Calendar — putting things that weren't appointments into appointment slots. What I did, how it went, what I made of it.
It worked, at first. At least the time that had passed was visible. Open a week and I could vaguely read back what I'd been living through.
It didn't last. A calendar is a forward-facing tool. It's where you decide in advance what happens when; it isn't built to accumulate what already happened. Cramming a day's worth of thinking into a one-line event title was off, and so was digging those entries back out later. Using a tool against its grain wears you down more than you'd expect. It's hard to keep going once you know you're forcing the fit.
What I actually needed was in between
After a while of that, it became clear what I was missing.
A to-do list only looks forward. It tells you what's urgent, not where it came from. A journal only looks back. It keeps what happened, but not how that connects to what you're about to do. And the parts that mattered most — why I decided to take this on, how it turned out, how I felt at the time, what I'd do differently next — had no home on either side.
These were always one continuous thing. Something happened, so a task appeared, I did it, it turned out a certain way, and I came away with a feeling about it. But the moment I picked up my tools, that line got chopped into pieces. The front went in the calendar, the middle in the to-do app, the end nowhere at all. Which meant reassembling the pieces was left to my memory — and human memory is spectacularly unsuited to that job.
I went looking. Note apps are free-form, and that freedom means no structure, so things just pile up over time. Project tools have plenty of structure, but no room for a person's day or how they felt about it. I tried running both side by side, and the moment there were two tools, the line broke again.
So I decided to build it myself.
One system that doesn't scatter
I set one rule going in: records have to connect to each other.
In Life OS, a record is one of six things. What you're actually trying to achieve (Initiative), what needs doing today (Task), a thought or a scrap of a meeting (Note), what happened that day and what state you were in (Check-in), a person and the context wrapped around them (People), and the conclusions you reach when you look back (Review).
The point isn't the six categories. It's that they hold onto each other. The task you wrote today belongs to an initiative, that initiative started from something that happened last week, a particular person is tangled up in it, and a month later a review pulls all of it back up at once.
One thing changed as a result: "why am I doing this again?" always has an answer waiting. I can reopen the list three weeks later and still know why that task is there. And once it's all cleared, something is left.
I decided not to have the AI summarize
Once this much existed, something else became possible.
Most productivity tools bolt on AI these days, and they mostly do the same thing: shorten long text, or tidy up a sentence. Useful enough, but it amounts to reorganizing what I just typed in. The tool still doesn't know me.
When the context sits in one place, the questions you can ask change. Why does this thing I've pushed back three times this month keep slipping. What the bad days had in common. How much time I've actually put into the thing I said would matter this year. None of that comes out of summarization. It requires tasks and days and people and goals to be sitting in the same place.
So the morning briefing, the feedback that comes a day later, and the monthly and annual reviews all rest on that. The AI doesn't get handed a document; it gets the context that's built up so far.
There's one thing I deliberately didn't do. No mechanical nagging — no "you've postponed this 3 times, stop postponing." Alerts assembled out of if-statements get ignored from the second time onward. Every judgment call is made by reading what's actually there.
These days I'm removing more than I'm adding
Most of what's needed is in. The six record types and the links between them, the daily briefing and feedback, the coach feed, monthly and annual reviews generated automatically. It runs in Korean, English, and Japanese, and installs on a phone like an app.
So the work now isn't adding features. I use it every day, and I keep taking things out.
There are screens that looked reasonable when I built them and that I then didn't open once in weeks. There were places that took two taps when there was no reason for two. There were options I added because they seemed nice to have, and all they did in practice was delay a decision. I find them one at a time and delete or merge them. Every time a feature goes, I can see the flow get a little smoother.
That's only possible because the person building it is the person using it daily. If you're watching usage through metrics alone, it's hard to know what's safe to remove. It isn't fast going, but whatever got in my way today can be gone this week.
I'm not claiming it's finished. Use anything daily and new friction keeps surfacing. But the direction is settled: I'm not building a tool that can do more things. I'm building one you stop noticing while you use it.
What I'll write here
This blog will cover what I've learned about managing yourself, how to actually use Life OS, and what happens while building it. Not only the decisions that worked out.
That's it for the first post.
If the tools you're using feel off somehow, I'd suggest one check before you go shopping for a new app. Pick any task off today's list and see whether you can immediately say why it's there. Then ask what's left after you clear it.
If neither answer comes easily, a different app probably won't change that.
Life OS keeps your tasks, initiatives, notes, days, people, and reviews connected as one record, and has AI read the whole of it to give you briefings and coaching. → life-os.cloud