Our story

We started Spellimus because we couldn't face another Friday list.

A teacher, an engineer, and a parent of three walked into a kitchen, looked at a crumpled A5 sheet of twenty words, and said "there has to be a better way to do this." There is.

2021

It started with a WhatsApp voice note.

Rania was teaching Year 4 in East London. Every Monday she'd send parents a voice note of the week's spelling list, because half her class had EAL parents who weren't sure how to pronounce "mischievous." It worked — attendance on Friday's test jumped — but it didn't scale past 30 kids.

She told her friend Oli, who is an engineer and a parent of three who'd just finished helping his youngest write "brocolly" for the sixth time. They spent a Sunday building a web page that played audio and let kids type back. His kids used it all week. His youngest got "broccoli" right on Friday.

2022

The pilot that convinced us.

We ran a six-week pilot at Oakfield Primary with three Year 5 classes. Practice completion went from 38% on paper lists to 72% on Spellimus. Teachers reported getting back about an hour a week. One parent told us her son had stopped saying he hated spelling. That was the moment it stopped being a side project.

We quit our jobs in July, moved into a cupboard-sized office in Hackney, and started calling schools.

2023

Three growing pains we're glad we had.

The streak incident. We launched streaks. Streaks worked beautifully on engagement and awfully on wellbeing — we had ten-year-olds in tears when the family went away for a weekend. We killed streaks two months in and redesigned around weekly practice rhythm, not daily guilt.

The voice that didn't sound like Mum. Our early TTS was good for American English and weird for everyone else. We worked with a phonics specialist in Manchester and recorded native UK voices, and we added the record-your-own-voice feature that EAL families had been asking for.

The tuition centre we didn't see coming. A 11+ tutoring chain in Wembley asked if they could assign lists to their Saturday cohorts. We hadn't thought about tuition centres at all. Groups became a first-class feature two months later.

2024

Growing up.

We're now a team of nine, in 40 schools and 12 tuition centres across the UK. We pick up about one new school a week, mostly from teachers telling other teachers. Rania still teaches on Fridays. Oli still pronounces his youngest's list at the kitchen table on Thursdays — though now the iPad does most of the heavy lifting.

We took a small investment from Ada Ventures and a group of angel teachers, on the condition that we stay weird about kids' screen time and never sell learner data. Both conditions feel easy to keep.

2025

What's next.

Arabic support (we have 2,000 children on the waitlist in the Gulf). Offline-first native apps. A proper phonics pack for early years. And — eventually — quiet, local, private on-device models so audio generation doesn't depend on an internet connection or anyone's API.

What we're not doing: AI tutors that chat with your child, gamification loops, any of the stuff that makes the rest of edtech feel like a casino. We'd rather stay boring and keep working.

How we decide things

Six rules we've never regretted.

When we're about to build something — or ship a feature — we run it through these. If it fails two of them, we don't ship.

01

The kitchen table test

If a tired parent at 7pm can't figure out what to do next without reading help text, it's broken. We design for Wednesday nights, not demos.

02

Teachers first, vendors last

Every feature we ship has been watched in a real classroom. We don't take feature requests from procurement; we take them from literacy leads.

03

No shiny things

Streaks died. Badges never shipped. No AI chat companions. No "levels." We believe in mastery, practice, and moving on — quickly.

04

Data stays with the child

Learner data is never sold, never shared for advertising, never used to train third-party models. If we ever change this, we'll close the service first.

05

Short is kind

Six to eight minutes a session. If a practice run goes over ten minutes, we count it as a product bug and fix it that week.

06

Sideways evidence

We measure mastered-word-retention, not just weekly minutes. Time-on-app is a terrible proxy for learning. We've got the graphs that prove it.

The team

A small team, one part-time dog.

Nine people in London, Lisbon and Leeds. We're hiring a Year-1–3 literacy advisor part-time — if that sounds like you, have a look.

RH

Rania H.

Co-founder · CEO

OW

Oli W.

Co-founder · CTO

PS

Priya S.

Literacy lead

MJ

Marcus J.

Head of design

AO

Amara O.

Teacher success

LG

Luis G.

Engineering

FK

Fatima K.

Arabic programme

🐕

Biscuit

Office dog

Try Spellimus free for 30 days.

For families, teachers, and tuition centres. No card, no setup call — just your first weekly list.

No card required · up to 6 children per family · cancel anytime