Cleo
Building credit and lending products.
Building credit and lending products.
A free iOS app that helps young adults get out of debt. Snap a photo of a credit card, student loan or car loan statement and the AI pulls out the balance, APR and minimum payment, then shows your total debt, which debt to pay first and your debt-free date. Cash AI is a voice and text finance coach that answers money questions, explains confusing bills in plain English and runs what-if scenarios before big decisions.
First product hire at a seed-stage fintech that rewards homeowners for everyday spending. Owned three core products from zero to scale: the Mesa Homeowner Credit Card, Mesa Mortgage and Mesa AI.
Director of Product and head of product for US Product and Innovation. I left to launch Mesa with my former engineering partner from Cash App.
Product lead for virtual cards and buy now, pay later, driving commerce and BNPL innovation at Cash App and Afterpay from zero to one.
Contract product and marketing leader helping the credit card startup (formerly CreditStacks) reach its $34M Series A, alongside $100M in receivables financing from WebBank. Jasper is a credit card platform backed by leading investors with over $248M in funding.
Led product for Denizen, the world's first global borderless bank account: one account for instant access to your money around the world.
Co-founded the personal safety app formerly known as SafeTrek, which became the #1 safety app on iOS and Android, raised over $7M in venture funding and went on to power safety features on Tinder. I had a financial exit from the company.
Led product innovation across TurboTax and Mint for three years, working weekly with Intuit founder Scott Cook for two of them.
Breathometer is a mobile platform that uses a small hardware device to detect a range of compounds from human breath. Its first product was a single-sensor breathalyzer for iOS and Android. The company raised over $30M from Shark Tank, Sir Richard Branson and top-tier VCs.
AgLocal was a two-sided platform that connected local family farmers with top chefs and restaurants throughout the USA. It raised $3.5M in seed funding from Andreessen Horowitz, and later became a consumer-facing subscription service.
Six US patent applications filed for Cash Balancer in March 2026. All are patent pending.
Multimodal AI pulls financial data from photographed receipts, paychecks, statements and loan documents, then validates it server-side, with on-device OCR as a fallback.
A system that uses multimodal AI to extract financial data from photographed documents (receipts, paychecks, credit card statements, loan documents), then applies server-side validation including APR swap detection, balance sanity checks, and range clamping. Extracted data populates type-specific data models across 7+ financial document types, with automatic fallback to on-device OCR when cloud processing is unavailable.
Turns a photographed bill or statement into a plain-language explanation read aloud by a consistent AI persona.
A system that converts photographed financial documents into plain-language voice explanations. A multimodal AI analyzes the document image, generates a conversational explanation using a consistent AI persona, and pipes the output to a text-to-speech engine for audio playback, enabling users to hear their bills, statements, and financial documents explained in simple terms.
Answers "what if" money questions against your real finances, with before-and-after impact on debt timelines, interest and cash flow.
A system that accepts natural language "what if" financial questions, ingests the user's real financial data, and uses an AI model with an allocated extended thinking budget to perform deep scenario analysis. Returns structured before-and-after comparisons showing projected impact on debt timelines, interest paid, and cash flow, along with actionable recommendations.
Reads investor emotion from voice check-ins, classifies behavioral finance biases and responds with adaptive spoken coaching.
A system that captures investor emotions through voice analysis during check-in sessions, extracts speech biomarkers (speech rate, pauses, duration), classifies behavioral finance biases (loss aversion, herd mentality, disposition effect, overconfidence, recency bias), integrates real-time portfolio and market intelligence, and generates adaptive AI coaching responses delivered via synthesized speech.
Anonymously aggregates voice-derived emotion across investors into a real-time crowd sentiment index that triggers proactive interventions.
A system that anonymously aggregates voice-derived emotion data from a population of individual investors to build a real-time crowd sentiment index, optimizes coaching strategies through a feedback loop that tracks effectiveness across different emotional states and market conditions, and triggers proactive interventions when crowd-level fear, FOMO, or anxiety exceed critical thresholds.
Gives licensed financial advisors voice-derived emotion updates across their client book, with outreach recommendations and consent controls.
A system that delivers voice-derived client emotion intelligence to licensed financial advisors and money managers, replacing the industry's quarterly check-in model with emotion-driven proactive engagement. Advisors receive real-time emotional state updates across their entire client book, AI-generated outreach recommendations with behavioral bias briefings and talking points, and configurable alerts triggered by client distress, disengagement, or crowd-relative emotional outliers, all governed by granular, revocable client consent.