Building Asermu: from idea to launch

The gap in the toolbox

I have spent years trying to write fiction and getting derailed by tools that almost worked. Google Docs is fine for a letter or an essay, but it crumbles under the weight of a novel with forty characters, three timelines, and a magic system that needs its own encyclopedia. Scrivener is powerful, but it carries the weight of two decades of accumulated features and a learning curve that makes you wonder if you should have just used a spreadsheet. The newer cloud-only tools felt slick, but something about storing an entire manuscript on someone else's server — with someone else's terms of service — never sat right with me. A novel is one of the most personal things a person can make. The tool that holds it should respect that. I never finished a novel in any of those tools. I told myself the tools were the problem. They were not the only problem, but they were enough of one.

So I started building something. Not out of confidence that I could do it better, but out of a stubborn refusal to keep working around tools that were almost right.

A name from before

The name took longer than the first prototype. I wanted something that was not an English compound word, not a clever portmanteau, not a Latin root dressed up in a tech suffix. I am Canarian. My ancestors, the Guanches, were Amazigh people who crossed to the islands long before anyone thought to write down why. The Amazigh languages carry words for writing and poetry that predate most of the languages software tends to be named in.

Asermu draws from that tradition. Asefru means poem or verse in Kabyle and across much of the Amazigh world. Aru means to write. The prefix ase- marks a tool or instrument — the thing that does the work. Asermu is not a documented word in any existing dictionary. It follows real Amazigh morphology, but it is new. A writing instrument, named in a language family that has been here since before the Latin alphabet reached North Africa.

That felt right for what I was building.

Offline-first, by design

The first real architectural decision was also the most consequential: everything runs in the browser. Your manuscripts, your characters, your worldbuilding notes — all of it lives in IndexedDB on your device, managed through Dexie.js. Nothing touches a server unless you explicitly choose to back up.

This was not the easy path. Next.js is built to talk to servers. Bending it into an offline-first application meant rethinking data flow, sync logic, and every assumption about where state lives. But the alternative was asking writers to trust that my startup would keep their servers running forever. I did not want to make that promise because I could not guarantee it. What I could guarantee was that your data stays yours, on your machine, accessible whether or not you have an internet connection. (I've written about this in more depth in Your novel shouldn't live on someone else's server.)

Cloud backup exists for writers who want it — encrypted, cross-device, opt-in. But the app works without it. That distinction matters.

Writers do not just write

Here is what surprised me most during development: the worldbuilding tools took more engineering effort than the manuscript editor itself. I expected the editor to be the hard part. TipTap made it manageable. What I did not expect was the sheer scope of what writers actually need alongside their prose.

A location system with nested maps and draggable pins. An encyclopedia with interlinked entries. A relationship graph that visualises how characters connect. A timeline with structured dates that can handle invented calendars. Character templates and name generators. A corkboard for rearranging scenes.

Writers do not just write. They build entire worlds, and they need tools that take that building seriously. Every module I added taught me something about how fiction gets made — not word by word, but layer by layer. The manuscript is the surface. Underneath it sits a structure that most readers never see and most writing tools barely acknowledge.

Warm by choice

Most software defaults to blue. Cool, professional, corporate blue. It signals trust and competence, which is fine for a bank. But a writing tool is not a bank. A writing tool should feel like a place where creative work happens.

Asermu uses warm earth tones — desert amber, terracotta, bone white. The UI draws on pintadera motifs, geometric stamps from pre-colonial Canarian pottery that the Guanches used in rituals and daily life. These are not decorative flourishes. They are a deliberate choice to make the tool feel like it was shaped by a specific person with a specific history, not assembled by a committee optimising for conversion metrics.

That extends to what the tool does not include. No tracking pixels. No data harvesting. Analytics run through Umami, which is cookieless and privacy-respecting. There is no writing assistant bolted on because writers do not need a machine finishing their sentences. They need a reliable place to do the work themselves.

The shape of what launched

Asermu launched in April 2026. The free tier is real — you get the manuscript editor, characters, locations, and timeline for a full project. Not a fourteen-day trial, not a feature-stripped demo. A working writing tool that you can use for as long as you want.

Paid plans unlock the deeper worldbuilding tools: plot grids, encyclopedia, relationship graphs, corkboards, multiple projects, export to PDF and EPUB. The pricing is straightforward. No hidden fees, no surprise rate increases.

I built this because I kept procrastinating on writing by trying every other writing app first. Eventually it seemed faster to build the one I actually wanted than to keep auditioning new ones. Asermu is not my first product, but it is the one I kept coming back to. I am launching it because the writers I never quite became should not have to keep auditioning tools either. If you are tired of tools that are either too simple to hold a novel or too complex to enjoy using, I would like you to try it.