More value from the cards you already carry.
Vane explores a simpler way to navigate card rewards, spending limits, and credit health through one considered mobile experience.
View working prototypeEveryday
•••• 2048
Best fit
3% estimated back
The problem
Too many rules for one everyday decision.
Managing several cards means remembering reward categories, tracking balances, and checking offer conditions. Vane brings those decisions into a clearer experience, helping users understand which card may suit their next purchase.
Reward rules are scattered.
Balances affect which card makes sense.
Benefits can be difficult to understand and track.
The product concept
Your cards. Working smarter.
Choose a card with context
Compare estimated rewards against projected balances and user-selected limits.
Understand the benefits
Separate confirmed cashback, pending benefits, and estimated points value.
Keep the wallet organized
Explore a unified view of cards, balances, reward rules, and offers.
Build healthier credit habits
Explore credit tracking, payment reminders, and balance-planning guidance without promising score improvements.
Product design showcase
One considered mobile experience.
All interfaces and data shown here are illustrative concepts. Cards, transactions, offers, scores, and charts are fictional.
Good morning
Your wallet
Benefits this month
$42.80
$18.20 confirmed · $24.60 pending
Sample Everyday
Estimated 3% back · within your limit
Recent activity
3 sample cards
Wallet
Everyday
•••• 2048
Travel
•••• 2048
38% of user-selected limit
Reward rule
3% groceries · active · cap applies
Illustrative only
Credit health
Simulated score
742
Illustrative history
Suggested actions
What is built
A working recommendation backend.
A working Python and FastAPI backend evaluates a proposed purchase against supplied card balances and reward rules.
22 passing automated tests, plus a verified local demo request
The backend does not include a database, user authentication, live bank connections, real credit-score data, or payment processing.
A recommendation example
A ranked answer with context.
Illustrative request
$120 grocery purchase · two fictional cards · current and pending balances supplied
Recommended: Sample Everyday
Ranks higher because its active grocery reward produces the strongest estimated value while the projected balance remains within the user’s selected utilization limit.
How the algorithm works
Evaluate constraints
Project balances, pending charges, reserves, limits, caps, and card eligibility.
Calculate value
Compare cashback and points through explicit, user-supplied point valuations.
Rank and explain
Return eligible options in order with a readable reason—or no eligible card.
AI opportunity · planned
Explain, don’t obscure.
A future AI layer could translate recommendation factors into clearer natural-language guidance, summarize changing reward rules, and help users explore “what if” purchase scenarios. Deterministic constraints and calculations would remain the source of truth.