A shared economy for chores nobody wants to do
A cross-platform app for assigning tasks and actually getting them done, through rewards and penalties.
Every chore app assumes the hard part is tracking. It isn't. The person handing out the task already knows what needs doing, and the person receiving it already knows they haven't done it. What's missing is a reason to care by Thursday. My bachelor work started from that: the interesting problem in delegated tasks is sustained motivation, and motivation is the part nobody budgets for.
Solo bachelor project at HEIA-FR, supervised by Pascal Bruegger. Requirements, survey, technology choices, both apps, the backend and the store submissions.
Two people, or one person alone, agree a shared economy: what the tasks are worth, what they cost to miss, and what the points can buy.
Every screen reads live from the database, scoped to the one active relationship, so both sides always see the same state.
A scheduled job closes the day: tasks left incomplete trigger their punishment, completed ones pay out their points.
Points buy the rewards the other person defined, which is the whole point: the incentive is negotiated, not imposed by the app.
A month grid per task, each day green, amber or red, so a pattern is visible rather than a single day's failure.
One codebase for both stores, chosen on a weighted comparison against React Native and Xamarin, not on taste
Identity, and the stream that the whole app's state hangs off
Eight collections, with security rules as the real access control: a document is reachable only by the two people on either end of it
The nightly reckoning, as an endpoint per time zone
The actual clock, firing each endpoint ten minutes after its own local midnight
instead of · Pairs only, like both apps I compared against
57% of survey respondents expected to hold both roles, and most wanted to assign tasks to themselves, a use case I hadn't foreseen. Rather than bolt on a solo mode, signup creates a relationship with yourself. Seven people answering eight questions changed the product more than a month of thinking had.
instead of · Strict midnight, or a lenient window up to the morning
The survey asked how strict the app should be and the answer was clear: strict, with a small margin for forgetting to tick something off. Midnight punishes the person who did the work and forgot to log it; 6 a.m. lets the deadline rot. Ten minutes is the smallest margin that respects both.
instead of · One function taking the time zone as a parameter
The platform binds a schedule to a deployed function, so making the job fire at local midnight in twenty-seven places meant twenty-seven endpoints. It is indefensible on any code-quality metric and it was the right call against the deadline. I wrote that in the report rather than hiding it. It is the clearest example I have of a decision that is bad by one measure and correct by another.
instead of · Presenting both halves as equally grounded
I could cite research for the reward mechanic. I could not for the punishment half, so I wrote “in our opinion” instead of dressing an instinct up as a finding. That asymmetry is still the most honest paragraph in the report.
Dutine tracks and manages tasks assigned to people, and motivates them to finish on time through a system of rewards and punishments rather than reminders.
Shipped on both stores from a single Flutter codebase, with a Node backend on Firebase and Google Cloud.