Together (Kopā)

Family chat
nortifications
food guide
Tasks
baby daily log
AI family planner
safety
Dashboard

Inspiration

Family life is not only about completing visible chores. It also includes planning, remembering, preparing, checking, coordinating, and taking responsibility for the final result.

A task such as “take the baby to the doctor” is actually a chain of smaller responsibilities: book the appointment, add it to the calendar, prepare documents, pack the baby’s bag, remember the time, travel to the appointment, and buy anything needed afterward.

Kopā, which means “Together” in Latvian, was created to make this invisible mental load visible and easier to share across the whole family.

What it does

Kopā is a collaborative Android family assistant for up to 15 family members.

The app combines:

Shared and recurring family tasks
Task assignment and workload visibility
A shared calendar
Shopping lists
A family group chat
Activity notifications with timestamps and unread counters
Baby meal planning and age-appropriate food suggestions
A baby photo gallery synchronized between parents
Daily and monthly family summaries
Location sharing and safety alerts
Family invitations through secure links
Battery-friendly synchronization and an eco/night mode

The new notification center shows who changed something, what was changed, the relevant section, and the exact time. Each section has its own unread badge, which is cleared when the family member opens that section.

AI-powered planning

Kopā uses the OpenAI API to turn a large or mentally demanding family responsibility into a structured, actionable plan.

For example, instead of creating one vague task called “Take the baby to the doctor,” the AI planner can generate:

Preparation tasks
Calendar events
Required shopping items
Suggested responsibility assignments
A practical sequence for completing everything

The generated result is returned as structured data that the user can review before adding it to the shared family plan.

The app clearly explains when AI is unavailable, when a free daily generation has already been used, or when a provider limit has been reached. It also supports an optional bring-your-own OpenAI API key flow. User-provided API keys are encrypted at rest and are never returned by the server.

How we built it

The Android application is written in Kotlin using Jetpack Compose and Material 3.

Room is used for local structured storage, while WorkManager handles background synchronization and scheduled work. The family relay is implemented in Node.js and deployed at ideaflow.ink.

The server acts as a lightweight coordination tower. It helps family devices discover each other, delivers encrypted synchronization data when direct communication is unavailable, and distributes signed APK updates.

For larger files, photos, video, live location, and safety streaming, the application can establish a direct WebRTC peer-to-peer connection. This reduces server storage requirements and keeps sensitive family media between participating devices whenever possible.

The application supports Latvian and English and can update itself securely through the family tower. Every release is verified using its version code, file size, APK structure, signature, and SHA-256 hash.

Challenges we ran into

One of the biggest challenges was reliable synchronization between multiple Android phones while keeping battery usage under control.

Fast chat delivery requires frequent communication, but continuous polling can drain the battery. Kopā therefore combines several strategies:

Direct P2P delivery when both phones are available
Relay-based fallback synchronization
Faster refresh while the chat is open
Slower refresh during normal usage
Eco mode
Optional night-time synchronization pause

Another challenge was designing a data model that could merge changes without losing completed tasks, chat delivery status, shopping updates, meal plans, or family activity history.

Building secure family invitations, update verification, encrypted API-key storage, and clear safety permissions also required careful attention.

Accomplishments that we are proud of

Built a working Android family collaboration application
Supports families of up to 15 members
Created automatic synchronization between multiple phones
Added direct WebRTC peer-to-peer communication
Built secure invitation links that open the Android application
Added a synchronized baby gallery and meal-planning workflow
Implemented an OpenAI-powered family planning assistant
Added a bring-your-own OpenAI API key option
Created a secure self-hosted APK update system
Added Latvian and English localization
Designed an activity history and per-section unread notification system

What we learned

The project demonstrated that family organization is not just a traditional to-do-list problem. A useful family assistant must understand dependencies, preparation, reminders, communication, safety, and the mental load behind each task.

We also learned how important graceful failure is. When AI, P2P, or the relay is temporarily unavailable, the app must explain what happened and continue working with locally stored data.

What's next for Kopā

The next steps are:

Improve conflict-free synchronization between many devices
Add richer AI-generated meal and family routine plans
Introduce optional encrypted backups
Improve accessibility and tablet layouts
Add more languages
Expand the notification center with filters and search
Complete production billing only if there is sufficient user demand
Publish the application through an official Android app store
Continue privacy and security testing

The long-term goal is to make Kopā a calm, private, and practical family operating system that reduces invisible mental load instead of creating more notifications and complexity.