I spent years building a finance tracker for myself. Last year I finally turned it into a product.

Most products start with a market. Mine started with annoyance.

I wanted to know where my money goes. Every finance app I tried demanded discipline: enter expenses by hand, configure endlessly, reconcile manually. The data entry always died after two weeks.

So I built my own — the first version was literally Perl + plain HTML. One principle drove every feature since: if input isn't automatic, the habit dies.

What "automatic" came to mean over the years:

- Bank webhook integration first (pay by card → expense appears in a second), then statement imports, then an AI engine that imports any bank statement without a dedicated parser

- Utility meters: send a photo to a Telegram/WhatsApp bot → AI reads the digits, detects which service the meter belongs to, applies the tariff

- Car costs: photo of fuel receipt + odometer → real cost per kilometer, fuel vs repairs vs maintenance separated

- Multi-currency with stored rates (unstable local currency makes single-currency analytics meaningless)

Last year I rebuilt it on a modern stack (FastAPI + Vue 3), added family groups, a tenant mode for landlords, localized it into 20+ languages with an AI translation pipeline, and opened it up: https://finman.vhworx.com or just message the bot: https://t.me/vikFinManBot / https://wa.me/4367764801396

Core features are free, premium covers AI quotas. It's early: the product is solid, distribution is the mountain I'm climbing now (SEO for a new domain is humbling — Google indexed the homepage and politely ignored the other 200 URLs for months).

Happy to share anything: the AI translation system, distribution lessons, or what it's like productizing a 100%-personal tool. The full story is here: https://finman.vhworx.com/story

Question for the community: for those who launched dev-tool-grade products to non-dev audiences — what distribution channel finally worked for you?