Veblen

Context

AI-powered market insights, insider trading signals, analyst recommendations, personalized stock recommendations, portfolio analysis and more, all in one elegant app.

What I did

Built a cross-platform codebase designed for both iOS and Android from a single source: a React Native client talking to a Node.js server with a Supabase Postgres database at the core, data pipelines, APIs and background jobs processing 70,000+ assets (stocks, ETFs, mutual funds, commodities, forex and crypto) across global markets. Integrated institutional-grade datasets: financial statements, earnings transcripts, analyst estimates, insider trading activity, ETF holdings and decades of historical data, with caching and real-time data feeds engineered for scale to reduce API costs and increase efficiency.

Across the app, a feature layer with an AI stock overview (buy rating, health score, risk score, market sentiment and a 1-month forecast), an 'Ask Why' explanation behind every insight, market signals (analyst ratings, price targets, insider trading, fair value, Piotroski score, bankruptcy risk), AI-simplified financials, news, daily hidden gems and daily movers, a 24/7 AI chat assistant, daily AI move recommendations, an AI tools suite (daily positions briefing, market overview, market news summary), OAuth sign-in with a personalized onboarding flow that surfaces AI stock picks that, on average, outperform 92% of investors historically, full position tracking and unlimited watchlists.

Designed the entire product in Figma, leading UX research, wireframing, prototyping, animations and the design system from concept to production, and led product strategy, software development, marketing and business operations.

Sourced international partnerships and negotiated data contracts with brokers and exchanges, from cold outbound emails to setting up meetings and negotiating the best data prices along the way.

Growth rested on organic content: short videos and slideshow posts edited in DaVinci Resolve and Affinity, plus UGC-style strategies built for Instagram, TikTok, Reddit and X.

Outcome

11M+ views, 380k+ likes, 160k+ shares and 13k+ followers across Instagram, TikTok, Reddit and X, driven by a predominantly international audience led by the US, Canada, UK, Australia and Singapore. Commercialization was ultimately constrained by the cost of institutional data licensing and exchange agreements, and I moved on.