Help people understand a loan before they commit to it.
Crypto-backed borrowing asks people to connect several unfamiliar ideas: collateral, networks, liquidation and fiat withdrawals. A simple-looking screen is not enough if the decision remains hard to understand.
What I owned
I shaped the product concept, UX strategy, identity, application flows and design-system foundations, taking the experience from an initial proposition to alpha.
The design tension
Keep the entry approachable while preserving the information people need to understand a financial commitment.
Inside the original workThe original case documents interviews, a problem-and-solution brief, onboarding, a loan calculator and complete repayment states. Aave Grants DAO’s October 2023 award report independently confirms the $25,000 grant.
02 / From understanding to decisions
The why. Then the interface.
A closer reading of the original work: the problem each choice addresses and the value it is intended to create.
01
Familiar entry, unfamiliar product
What the work shows
The source brief identifies complexity and low awareness as barriers to borrowing.
The design response
The proposal uses familiar Google/email entry points and introduces the product before the detailed loan choices.
Why it matters
Help a newcomer reach the product without needing to learn every crypto concept at once.
02
The amount is only part of the decision
What the work shows
The calculator connects supplied ETH to requested USD, fees, an estimated liquidation price and loan terms.
The design response
Keep these details beside the amount, then create a separate review moment.
Why it matters
Help the person compare the consequences of a choice before confirming it.
03
A loan creates an ongoing task
What the work shows
Funding and repayment involve network selection, alternative payment routes and processing states.
The design response
Design the return journey: amount → review → processing → confirmation, with an explicit network acknowledgement.
Why it matters
Make the product useful when the person needs to manage or close the loan.
03 / Connecting the experience
More than a conversion funnel.
Select a stage to explore its purpose.
01
The moment
Understand the product and complete onboarding.
The design intention
Establish a recognizable and consistent introduction.
Portfolio journey map, reconstructed from the published onboarding, wallet, calculator and repayment screens.
04 / The interface, up close
One journey. Every important moment.
Individual screens from the original application, reframed around the decisions they support.
01 / Original interface
Balances, loan health and next actions
Pairs supplied assets and borrowed balance with liquidation price, loan health and a next action.
Design intentionKeep the loan state, relevant details and next action connected.
A connected concept-to-alpha product spanning identity, onboarding, borrowing and repayment. Ambos Finance received a $25,000 Aave Grants DAO grant to develop its proof of concept.
For the person
A clearer relationship between the amount borrowed, the collateral and the next action.
For the product
A coherent lending journey that continues after the first transaction.
For delivery
Reusable interface patterns and explicit states for implementation.
A tradeoff worth keeping visible
Fewer steps can feel faster, but financial details still need attention. I would keep essential conditions visible and make the review step deliberate.
Proposed validation / Next iteration
Ask someone to choose an amount and explain the collateral risk in their own words.
Observe whether they can find the fee, terms and repayment route without help.
Compare comprehension and completion errors before measuring conversion.
Evidence & presentation notes +
The alpha milestone is documented in the newer career material. A measured improvement in comprehension or conversion is not established.
These explanations and presentation layouts are newly prepared from the available original artifacts. The validation tasks are proposals, not research already completed.
The research artifacts are self-reported in the original project presentation. Exact work dates need confirmation before publication, so this local preview uses the broad year only.