Zolvo (Backed by Y Combinator)
AI SaaS that automates servicing for commercial lenders.
00

problem
Commercial lenders operate on 1990s software. Teams spend 6+ hrs/day matching wire payments manually, with 16–20 errors/month and 3-day processing cycles. Every new client means more headcount, not better software.
solution
Zolvo sits as the AI layer between bank accounts, loan management systems, and ops teams, handling reconciliation and collections automatically. Your team reviews only the exceptions.
As the founding designer at a YC Spring 26 startup, I owned the full product experience from 0→1.
the goal: make complex back-office operations feel simple, clear, and trustworthy at a glance.
Impact:
-Created a Design System
-Designed 5 core product modules from 0→1 in under 8 weeks
-Landing page launched
-Validated flows through 10+ live walkthroughs
-Designed toward clear KPIs: time-to-first-action, exception review rate, and reconciliation error reduction
Research:
I interviewed ops leads and commercial lenders to map their real workflows. The core insight: the biggest pain wasn't any single task. It was the cognitive load of tracking everything at once with no single source of truth.

Design System:
Before designing any screens, I established a visual foundation that felt like an upgrade to what these teams already knew, not a replacement. Dark and Light base, high-contrast, clean utilitarian typography, and dense consistent components, all chosen to feel modern and precise while staying familiar to ops teams used to legacy software.


Design and fast iteration:
Then I translated the highest-friction points into four product pillars: Invoice Verification, Real-time Reconciliation, Loan Collections, and Portfolio Monitoring. Every decision was anchored to one question: does this reduce what an ops person has to hold in their head?
Multiple directions were scrapped along the way. The first reconciliation view showed all transactions at once, creating the exact overwhelm we were solving for. The final design surfaces only unmatched items by default.
Invoice Verification:
I introduced confidence scoring and evidence-based alerts so operators could see why the system flagged something, not just that it did. This built trust in the automation and reduced unnecessary manual overrides.

Real-time Reconciliation:
I designed a quarantine state for unmatched transactions so nothing gets lost or silently dropped. The default view shows only items needing attention; clean matches are archived automatically.

Loan Collections:
I designed a pipeline view mapping each account to its current collections state, with one-click access to communication history and next recommended action. A process that lived in spreadsheets became a single managed workflow.

Communications Hub:
I designed a unified timeline aggregating all borrower outreach across email, phone, and portals, so any team member can see the full history without digging through inboxes.

Portfolio Monitoring:
Designed for a 30-second daily check-in. Key risk signals surface at the top; color and iconography signal urgency without creating noise.

What's next:
Next priorities include a mobile-first exception review experience and a client onboarding flow. Once live, primary success metrics will be reconciliation error rate, time in platform per day, and exception resolution time.
year
Mar. 2026 - Present
interface
Desktop
tools
Figma, Cursor
category
UI/UX
01

02

03


