How We Built TarsierAlpha: The Story Behind the Scanner

Beginner 8 min read Tarsier Alpha

How We Built TarsierAlpha

TarsierAlpha started as a personal frustration.

The problem wasn't knowledge. After years of trading options using Gap Fill plays, Oversold Bounce setups, and Catalyst Plays — the three strategies that produced results like MEI +75.71%, NFLX +197%, APH +167%, and MSFT +$1,122 in a single day — the methodology was proven. The problem was time.

Finding high-quality setups manually meant spending 2–3 hours every evening scanning hundreds of charts, checking RSI levels, looking for gap patterns, evaluating options chains. For a working professional with capital to deploy and a proven strategy, that time cost was the barrier.

The question became: can a scanner do what 3 hours of manual scanning does — but in minutes, automatically, every single day?

The Build Philosophy: Radical Transparency

Before writing a single line of code, one principle was established that shaped everything else: radical transparency.

Most trading products hide their losses. They show you cherry-picked wins. They display returns that don't account for commissions, slippage, or failed trades. The industry's marketing culture is built on showing only the best.

TarsierAlpha would do the opposite. Every trade — wins and losses — would be shown publicly. The -98% HOOD loss happened. It's in the record. The week-one -60% drawdown happened. It's in the record. The recovery to +25% overall is meaningful specifically because the losses are shown alongside it.

This isn't just ethics — it's strategy. In a market saturated with gurus flashing Lamborghinis and showing only their wins, a platform that shows everything builds a fundamentally different kind of trust.

The Technical Architecture

TarsierAlpha is built as a Flask/Python web application on a PostgreSQL database, scanning 500–8,000 tickers daily using data from yfinance, Financial Modeling Prep (FMP), Finnhub, and Tradier.

Every morning, the scanner runs through its universe of tickers and evaluates each against three strategy modules:

Gap Fill Module:

Oversold Bounce Module:

Catalyst Play Module:

Each candidate that passes strategy filters receives a composite Entry Score from 0–100, aggregating signal strength across technical and fundamental dimensions. Only scores of 62+ appear in the Top Candidates section. Scores below 62 feed the Watchlist as "approaching setup" notifications.

The Paper Trading Validation Phase

Before opening to paid subscribers, TarsierAlpha entered paper trading validation — tracking every platform-flagged setup as a real trade with real options prices, real premiums, and real exit prices.

The first 10 trades produced a 50% win rate with the following notable results:

The week-one drawdown reached -60% before recovering to +25% by week three. Showing this trajectory — the loss, the recovery, the methodology holding through adversity — is the entire point of the paper trading phase.

What's Coming Next

TarsierAlpha is actively developing:

The name "Tarsier" comes from the tarsier — a nocturnal primate with enormous eyes relative to its body, capable of seeing in near-complete darkness. The eyes see what others miss. That's the product.

See the platform live: tarsieralpha.com | Related: How TarsierAlpha's Entry Score Works

Ready to see real setups?

TarsierAlpha scans 500+ stocks nightly and surfaces the best options trades automatically.

Start Free Trial Join Free Alerts