Case study · Built by Medellis

Overpriced.ai — the growth rate hiding inside every stock price

Reverse-DCF analysis that turns "is this expensive?" into a number you can argue with — plus the screener, alerts, and bot that make it a habit.

Verticalfinancial-services / investment-analysis
Statuslive
See itoverpriced.ai

01The problem

Most equity research hands you a price target and hides the assumptions. Expectations investing flips that: start from today's price and extract the growth it implies, then ask whether the business can plausibly deliver it. The method — popularized by valuation teaching like Damodaran's — is powerful and tedious in equal measure, which is why almost nobody does it by hand for more than one ticker.

02What we built

03How it's built

The valuation math is deterministic code with the model layered on top for explanation — the reverse of most "AI finance" products, and the reason outputs are reproducible. Scheduled jobs refresh the screener and snapshots daily.

stack — Claude grounded commentary · live market data feed · Supabase data & auth · Stripe tiers · Telegram Bot API · Vercel crons · Android / Compose
Rule of the product: the AI explains numbers the engine computed. It never invents them.

04Where it stands

Live at overpriced.ai with subscriptions running across web, Telegram, and Android. It's our standing demonstration that agentic features — bots, scheduled analysis, alerting — belong on top of rigorous, deterministic cores.

Overpriced.ai is an analysis tool, not investment advice.

Want AI that shows its work on top of your numbers?

Deterministic core, explainable AI layer — the pattern transfers to pricing, risk, and ops. A 30-minute fit call tells you if it fits yours.

Book a fit call