prospect-terminal

Front Page Terminal
Backtest Portal
Specific Stock: AI Overview Post Scrape
Specific Stock: Direct information that has been pulled

Inspiration

I was inspired by how fragmented stock research still is.

Investors constantly jump between price charts, SEC filings, financial news, macro dashboards, and social sentiment just to form one view on a stock. Most tools either show raw information without synthesis, or generate signals without clearly explaining where they came from.

I wanted to build something that feels more like a real market workstation — a product that can pull together live market context, explain the setup behind a stock, and let users test whether that setup actually works over time.

That became prospect-terminal.

What it does

prospect-terminal is an explainable market intelligence platform for stock research and strategy testing. It combines:

live stock prices real SEC 10-Q and 10-K filings provider-backed financial news macro context social sentiment from X and Reddit

From those inputs, it generates:

a Prospect Score a directional thesis like bullish / neutral / bearish an explainable score breakdown filing-driven catalysts and risks a historical score path and a backtestable signal

The goal is to help users move from fragmented market data to a clear, defendable stock view.

How we built it

I built prospect-terminal with:

Next.js TypeScript Tailwind CSS FastAPI Python MongoDB

On the data side, I used:

live quote ingestion SEC EDGAR APIs for 10-Q and 10-K filings provider-backed financial news social sentiment integration from X and Reddit historical score generation for backtesting

On the backend, I built services for:

quotes filings news macro context social sentiment scoring backtesting

I also used MongoDB as a cache/persistence layer so the app could stay reliable even when live sources were slow or failed.

Challenges we ran into

One of the hardest parts was balancing live data realism with demo reliability.

I had to solve for:

getting multiple data sources to work together cleanly keeping the product resilient when providers fail turning SEC filings into something actionable instead of just raw text making the score explainable instead of black-box improving historical score generation so the backtest felt more credible cleaning up leaderboard logic so it reflected real setups instead of junk or stale test symbols keeping the frontend polished without breaking the styling pipeline

A big challenge was making the product feel like a serious market workstation instead of just another hackathon dashboard.

Accomplishments that we're proud of

I’m proud that prospect-terminal is not just a pretty dashboard.

It now includes:

live stock prices real SEC filing ingestion provider-backed news social sentiment from X and Reddit explainable scoring historical signal generation backtesting against buy-and-hold a polished terminal-style UI

I’m especially proud that the product can both explain a setup and test that setup historically in the same workflow.

What we learned

I learned that building useful financial software is not just about getting access to more data — it’s about deciding:

which signals matter how to synthesize them how to keep the model interpretable and how to present the output in a way users can actually trust

I also learned how important fallback design, data normalization, and UI discipline are when working with multiple live data sources under hackathon constraints.

What's next for prospect-terminal

Next, I want to:

expand the tracked stock universe improve market-relative ranking strengthen the weighting of the social signal layer deepen historical signal coverage add alerts and watchlist workflows improve the strategy engine continue pushing prospect-terminal toward a real decision-support terminal for active investors

The long-term goal is to turn prospect-terminal into a platform that helps users move from noise to conviction with signals they can actually defend.