Case studies
For now, these are the products we built and run ourselves. Each follows the same path we use with clients: problem, approach, solution, outcome.
Client case studies are coming. We publish them only once the client has approved the write-up. If you would like to know more about client work, get in touch.
Kepton: private reply drafting on the phone
Problem
People want help writing replies in busy chats, but most assistants send private conversations to a server and need an account.
Approach
Keep everything on the device. Run the model locally, learn the user's own tone, and ask for nothing beyond a one-time model download.
Solution
Kepton is an Android app that drafts replies in your own voice and shows conversation pattern insights. It needs no account. See its privacy policy and terms.
Outcome
Shipped and live on Google Play. It proves we can take an on-device AI idea from concept to a store release.
Webonika Fitness: a white-label app for gyms
Problem
Gyms struggle to keep members engaged between visits, and trainers cannot personalise plans for every member.
Approach
Give each gym an app under its own brand, and let AI handle the repetitive planning so trainers can focus on coaching.
Solution
Webonika Fitness lets members scan a body report and receive an AI diet and workout plan, track their day, and compete on leaderboards, with tools for trainers. See the fitness industry page for how gyms use it.
Outcome
Live as a product that gyms can publish under their own brand. We learn from real gym use how to improve plans and tracking.
Webonika for Schools: in development
Problem
Teachers spend hours on quizzes, worksheets, papers and report cards, and on repetitive parent communication.
Approach
Generate drafts from the teacher's own material, keep the teacher in control of what is approved, and begin with coaching institutes.
Solution
Webonika for Schools is an AI layer for institutes and schools that produces quiz, worksheet and report drafts and automates parent messages.
Outcome
Not yet launched. We are running early access with institutes. Join the early-access list.
What these show
The same method sits behind client work: find the one workflow that matters, prototype on real data, ship, then improve. Explore our services or how we price.
Have a process worth automating?
Tell us the one task eating your team's time, and we'll show you what AI does with it. We reply within one working day.