Before the public launch, the bot spent several years as a closed internal tool for one team with a daily media flow. This page explains why Flowmark is built the way it is, and shows the numbers that tested the approach.
Deciding to add a logo takes a second. But the actual route is longer: download the file, open the editor, find and adjust the asset, check the proportions, select an export preset, wait for the render, and send the result back. When the task is done once, that is fine. When there are dozens of such tasks a day, it starts to feel like the main job.
Anything that repeats every day should take one click. Everything that changes rarely — assets, logo versions, caption presets, processing parameters — should live outside the daily route and stay out of the way. This creates the split: quick actions remain in the chat, settings go to the Mini App.
The closed version of Flowmark processed tens of thousands of files before the service became public. These numbers show which scenario became the core use case — and why the public version is organized around it. Current Flowmark statistics are in the “Proof” section on the home page.
In one year: 7,243 logo operations, 7,517 watermark operations, 551 captions, 337 face masking operations, and 455 video compression operations. The core scenario is quick branding for a steady flow; everything else was added as adjacent tasks grew out of this route.
Chat — for quick actions: send a file, press a button, get the result. The Mini App — for settings: assets, versions, presets, captions, face masking, video parameters, team members. The brand kit is portable: it can be exported as a .flowpack, moved to another account or handed off to another team.
The idea stayed the same: the path should be faster and easier when you repeat it every day.
