XISLABS

Case study · Mobile app · Pre-launch

LifeEchos: a voice-first life-recording app for Android, built from idea to Google Play testing

LifeEchos lets people capture moments, ideas and memories by speaking. They keep journals and diaries, ask an AI about their own records, organise tasks and store sensitive information in an encrypted vault for their family. We built the app and its backend. Version 1.0.0 is in Google Play internal testing.

Echo IntelligenceWhen did I last servicethe car?On 14 March. You notednew brakes were fittedat the same visit.Sources1Car service doneJournal · 14 March2Brakes replacedJournal · 14 March
Illustration
Client
The founder of LifeEchos
Country
United States
Product
LifeEchos, an Android app
Engagement
May to November 2026
What we built
The Android app, the backend, the AI features and the encrypted vault
Status
Pre-launch. Version 1.0.0 is in Google Play internal testing. Store billing is not connected yet
  • 90+app screens
  • 90+voice commands
  • 11drawers in the Legacy Vault
  • 5custom Android modules

The problem

  • Moments are lost because typing is slow. By the time someone has opened a notes app, the thought has gone.
  • Life is scattered across apps. Notes, journals, tasks, finances and documents all live in different places.
  • Records are hard to find later. A memory is only useful if it can be found again.
  • Families inherit chaos. Accounts, documents and memories are rarely organised for the people who will need them.

What we built

Hands-free capture.

The user says “Echo” and speaks. The app understands more than 90 voice commands, takes live dictation and works with the camera, calls and texts. A home-screen widget adds a voice button, quick typing and recent activity.

EchoListeningDictationPick up flowers forMum on FridaySay “Echo” and speak
Illustration

Writing spaces.

Notes, a journal, a diary and reflections, with titles written by AI. Private Spaces lock behind a PIN or biometrics.

An AI that answers from the user's own records.

Echo Intelligence answers questions about what the user has recorded and shows its sources. A separate assistant gives general spoken answers.

Echo IntelligenceWhen did I last servicethe car?On 14 March. You notednew brakes were fittedat the same visit.Sources1Car service doneJournal · 14 March2Brakes replacedJournal · 14 March
Illustration

The Legacy Vault.

Entry needs two factors: a vault PIN plus a fingerprint or face. It has 11 drawers, encrypted files and three access levels, so a user decides exactly what a trusted person can see.

Legacy Vault1PIN2FingerprintBankingLoginsPropertyInsuranceMedicalDocuments+ 5 more drawersLocks after one minute
Illustration

Life tools and sharing.

Finance with receipt reading, tasks, a calendar, call routines and a family tree. Banking items, logins and property records can be shared through encrypted links that expire or are view-only.

From a spoken thought to an answer later

5 steps. Use the arrows to go through them.

How private data is protected

  • Two factors for the vault. A PIN and a fingerprint or face, every time.
  • It locks itself. The vault locks after one minute without use. Private Spaces lock after 30 seconds, one minute or five minutes, as the user chooses.
  • Files are encrypted. So are the links used to share them, which can expire.
  • AI runs on the server only. The app itself holds no AI keys.
  • Actions need a yes. When the AI proposes an action, the user has ten minutes to confirm it.

Where it stands

Built and in the appNot yet available
Voice capture, journals, the vault, Echo AI, finance and the widgetAn iOS version
Membership rules, built and tested on the serverReal payments: store billing is not connected yet
Version 1.0.0 in Google Play internal testingA public release on Google Play
Speaker verification by voice

Status as of November 2026.

Under the hood: stack

App: React Native with Expo and TypeScript, with five custom Android modules written in Kotlin for the widget, vault encryption, voice print, PDF pages and sharing.

Backend: Supabase for the database, sign-in, file storage and server functions, with 11 server functions and 35 database migrations.

AI: large language models from one provider, with a second provider as a fallback, called only from the server.

Speech: on-device speech recognition and text to speech.

Have an app idea that needs to reach the store?

Tell us what it should do on a 30-minute call. You leave with a plan and a cost, in writing.

Book a discovery call