All case studies
Project

Sogon-Sogon: a voice journal that gives a psychiatrist the weeks between appointments

Patients with depression see their psychiatrist for minutes at a time, and the weeks in between go unrecorded. Sogon-Sogon is a voice-first AI journal that turns a patient’s day into a diary entry for them and a report their clinician can read.

At a glance

  • A conversational AI, "Sogoni", that listens by voice and writes the day’s diary entry from the conversation
  • Emotion analysis over every entry, summarised into a report and a dashboard for the patient’s clinician
  • A peer community and a location-based clinic finder around the journal, with partner clinics managed from an admin
  • Planned, designed and shipped in two weeks; placed fourth at the ToBeDoctor startup competition
Two screens of the app: Sogoni asking how the day was with five mood faces, and Sogoni listening while the patient speaks
Sogoni asks how the day went, then listens.

The problem

Treatment for depression runs on what a patient can remember in a short appointment. How the last three weeks actually went, which days were bad and what happened on them, is mostly lost by the time they sit down, and the psychiatrist is left working from a summary the patient gives under pressure.

Keeping a journal would fix that, but writing one is exactly the kind of effort depression takes away. The record that would help most is the one least likely to exist.

What we built

A mobile app built around one conversation a day, and three services around it:

Sogoni, the journal
The patient talks, by voice or text, to a language-model companion that asks about the day and listens. The conversation becomes a written diary entry, so nothing has to be typed.
The clinician’s report
Each entry is analysed for emotion, and the entries are summarised into a report and a dashboard the patient’s psychiatrist can review before an appointment.
Woongsung-Woongsung, the community
Patients can share entries and read other people’s, and a board collects information and first-hand reviews of medications.
Bangbang-Gokgok, the clinic finder
A location-based search for psychiatric clinics nearby, with reviews, and an admin for managing partner clinics.

Two things shaped every screen. The first was effort: one tap to start talking, a mood picked from five faces, and a soft, quiet interface rather than a medical one. The second was trust: diaries and medical information are the most private data a person has, so how it is stored and who can see it was designed in from the start.

How it was built

Two weeks from the first requirement to release. The first week covered planning, the interface and prototype in Figma, and the conversational AI; the second, the database, the clinician’s admin pages, testing and release. The app runs on iOS and Android from one build.

It was entered in the ToBeDoctor startup competition, with the pitch below, and placed fourth.

The pitch deck as a grid of slides: the problem, the patient and clinician journeys, the solution and the app screens (opens full size in a new tab)
The competition pitch: the problem, the patient’s and the clinician’s journeys, and the app.

More reading

Case study

How a franchise network let buyers search 100,000 products with a photo

Buyers on a promotional-products franchise network knew what they wanted but not what it was called, and keyword search across more than 100,000 products could not help them. An AI search now finds products from a photo, a phrase, or both, on the head-office site and every franchise storefront.